AMD EPYC 7002 / 7003 series NAS (32 GB) vs Intel Arc B580 for local AI
Pick two to four systems or devices to see which runs more models, faster, for the kind of work you do.
Winner: Intel Arc B580
For general chat: It runs 14 more LLMs (477 vs 463).
 AMD EPYC 7002 / 7003 series NAS (32 GB)
|
 Intel Arc B580
|
|
|---|---|---|
| Memory | 32.0 GB unified | 12.0 GB VRAM |
| Usable for models | 26.2 GB | 12.0 GB |
| Memory bandwidth | 205 GB/s | 456 GB/s |
| FP16 compute | 1.0 TFLOPS | 27.3 TFLOPS |
| Launch price | — | $249 |
| Released | 2019 | 2024 |
| LLMs that run | 463 of 619 | 477 of 619 |
| LLMs that run well | 395 | 425 |
| Image models that run | 57 of 60 | 54 of 60 |
| Video models that run | 37 of 43 | 14 of 43 |
| Decision models that run | 67 of 70 | 68 of 70 |
| Top-10 LLM average score | 98 | 92 |
| Top-10 LLM average speed | 39.5 tok/s | 39.8 tok/s |
| Best LLM | Qwen3.6 35B-A3B | K2 Horizon 7B |
Highlighted values are the best of the two. Model results assume 8K of context and the default macOS GPU memory limit on Macs.
LLMs
463 of 619 LLMs run on the AMD EPYC 7002 / 7003 series NAS (32 GB). 26.2 GB usable for models.
477 of 619 LLMs run on the Intel Arc B580. 12.0 GB usable for models.
| Model | QualityEPYC 7002 / 7003 series NAS | SpeedEPYC 7002 / 7003 series NAS | RunsEPYC 7002 / 7003 series NAS | QualityArc B580 | SpeedArc B580 | RunsArc B580 |
|---|---|---|---|---|---|---|
| Qwen3.6 35B-A3B | 88Excellent quality (A+) · 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 82.6 tok/s | Runs welloffloads to RAM |
| MiniCPM-V 4.7 35B-A3B | 88Excellent quality (A+) · 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 84.6 tok/s | Runs welloffloads to RAM |
| Qwen3.5 35B-A3B | 88Excellent quality (A+) · 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 82.6 tok/s | Runs welloffloads to RAM |
| Nex N2 Mini (35B-A3B) | 87Excellent quality (A+) · 87 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 84.9 tok/s | Runs welloffloads to RAM |
| Gemma 4 26B-A4B | 87Excellent quality (A+) · 87 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. | 32.7 tok/s | Runs great | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 117 tok/s | Runs welltight fit |
| Sage-A1.5 35B-A3B | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 82.6 tok/s | Runs welloffloads to RAM |
| GLM-4.7 Flash (30B-A3B) | 83Great quality (A) · 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 95.4 tok/s | Runs welloffloads to RAM |
| Ornith 1.5 35B-A3B | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 82.6 tok/s | Runs welloffloads to RAM |
| Apodex 1.1 Mini (35B-A3B) | 84Great quality (A) · 84 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 82.6 tok/s | Runs welloffloads to RAM |
| Ornith 1.0 35B-A3B | 84Great quality (A) · 84 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 82.6 tok/s | Runs welloffloads to RAM |
| Agents-A1 35B-A3B | 82Great quality (A) · 82 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 84.9 tok/s | Runs welloffloads to RAM |
| Qwen AgentWorld 35B-A3B | 82Great quality (A) · 82 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 86.2 tok/s | Runs welloffloads to RAM |
| Raon-1 mini Preview (30B-A4B) | 83Great quality (A) · 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking. | 29.3 tok/s | Runs well | 73Great quality (A) · 73 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 53.0 tok/s | Runs welloffloads to RAM |
| Laguna XS 2.1 (33B-A3B) | 77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 78.8 tok/s | Runs welloffloads to RAM |
| Nemotron Cascade 2 30B-A3B | 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. | 35.2 tok/s | Runs great | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 88.9 tok/s | Runs welloffloads to RAM |
| Flint (40B-A2.5B) | 75Great quality (A) · 75 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. | 58.5 tok/s | Runs welltight fit | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 43.2 tok/s | Slowheavy RAM offload |
| KAT-Coder V2.5 Dev (35B-A3B) | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 86.2 tok/s | Runs welloffloads to RAM |
| Qwen3 30B-A3B Instruct 2507 | 73Great quality (A) · 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs great | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 87.5 tok/s | Runs welloffloads to RAM |
| LFM2.5 24B-A2B | 71Great quality (A) · 71 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 40.4 tok/s | Runs great | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 151 tok/s | Runs welltight fit |
| Nemotron 3 Nano 30B-A3B | 73Great quality (A) · 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 35.2 tok/s | Runs great | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 88.9 tok/s | Runs welloffloads to RAM |
| DiffusionGemma 26B-A4B IT | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 31.2 tok/s | Runs great | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 113 tok/s | Runs welltight fit |
| Hunyuan Hy-Memory 30B-A3B | 80Great quality (A) · 80 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 59Good quality (B) · 59 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 95.8 tok/s | Runs welloffloads to RAM |
| Qwen3.8 27B | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 tok/s | Slow | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs welloffloads to RAM |
| Qwen3 30B-A3B | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs great | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 87.6 tok/s | Runs welloffloads to RAM |
| Qwen3-VL 30B-A3B Thinking | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs welltight fit | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 85.4 tok/s | Runs welloffloads to RAM |
| Nemotron 3.5 Lightning 30B-A3B | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 99.1 tok/s | Runs welloffloads to RAM |
| Qwen3.6 27B | 89Excellent quality (A+) · 89 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 tok/s | Slow | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs welloffloads to RAM |
| Qwen3.5 27B | 88Excellent quality (A+) · 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 tok/s | Slow | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs welloffloads to RAM |
| Qwen3-VL 30B-A3B Instruct | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs welltight fit | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 85.4 tok/s | Runs welloffloads to RAM |
| Nanbeige4.3 33B | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 4.2 tok/s | Slowtight fit | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 3.9 tok/s | Slowheavy RAM offload |
| Sarvam 30B | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 41.2 tok/s | Runs welltight fit | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 111 tok/s | Runs welloffloads to RAM |
| Muse Glimmer 30B | 87Excellent quality (A+) · 87 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 4.7 tok/s | Slow | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 16.5 tok/s | Runs welloffloads to RAM |
| North Mini Code 1.0 (30B-A3B) | 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 67Good quality (B) · 67 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 64.6 tok/s | Runs welloffloads to RAM |
| gpt-oss 20B | 70Great quality (A) · 70 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 36.4 tok/s | Runs great | 70Great quality (A) · 70 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 66.9 tok/s | Runs welloffloads to RAM |
| TMax 27B | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 tok/s | Slow | 75Great quality (A) · 75 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 10.1 tok/s | Runs welloffloads to RAM |
| Qwen3-Omni 30B-A3B Instruct | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs welltight fit | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 73.5 tok/s | Runs welloffloads to RAM |
| Trinity Mini (26B-A3B) | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 135 tok/s | Runs welltight fit |
| Fara 1.5 27B | 83Great quality (A) · 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 tok/s | Slow | 73Great quality (A) · 73 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 10.1 tok/s | Runs welloffloads to RAM |
| Gemma 4 31B | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 4.5 tok/s | Slowtight fit | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 10.6 tok/s | Runs welloffloads to RAM |
| Qwen3 30B-A3B Thinking 2507 | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs great | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 87.5 tok/s | Runs welloffloads to RAM |
| Qwen3 Coder 30B-A3B | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 tok/s | Runs great | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 87.5 tok/s | Runs welloffloads to RAM |
| UI-Mate 27B | 81Great quality (A) · 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 tok/s | Slow | 72Great quality (A) · 72 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 10.1 tok/s | Runs welloffloads to RAM |
| ERNIE 4.5 21B-A3B | 60Good quality (B) · 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 135 tok/s | Runs welltight fit |
| Maple Preview (20B-A1B) | 69Good quality (B) · 69 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 52.7 tok/s | Runs great | 58Good quality (B) · 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 168 tok/s | Runs welltight fit |
| Axion 1.0 35B | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 84.9 tok/s | Runs welloffloads to RAM |
| LFM2 24B-A2B | 54Fair quality (C) · 54 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 tok/s | Runs great | 39Basic quality (D) · 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 156 tok/s | Runs welltight fit |
| ERNIE 4.5 21B-A3B Thinking | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 135 tok/s | Runs welltight fit |
| NorthCUA Mini | 56Good quality (B) · 56 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs welltight fit | 41Fair quality (C) · 41 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 84.9 tok/s | Runs welloffloads to RAM |
| EXAONE 4.5 33B | 77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 tok/s | Slowtight fit | 77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 3.9 tok/s | Slowheavy RAM offload |
| Ling Mini 2.0 (16B-A1.4B) | 52Fair quality (C) · 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 42.6 tok/s | Runs great | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 151 tok/s | Runs welltight fit |
| Devstral Small 2 (24B) | 73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 tok/s | Slow | 63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 12.6 tok/s | Runs welloffloads to RAM |
| Gemma 4 12B | 78Great quality (A) · 78 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 11.7 tok/s | Runs well | 78Great quality (A) · 78 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. | 35.8 tok/s | Runs great |
| Groot 19B A3B 261023 | 54Fair quality (C) · 54 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 47Fair quality (C) · 47 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 115 tok/s | Runs welltight fit |
| Kimi Linear 48B-A3B | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 61.7 tok/s | Runs welltight fit | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 34.0 tok/s | Slowheavy RAM offload |
| Qwen3-VL 32B Instruct | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 4.2 tok/s | Slowtight fit | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 3.7 tok/s | Slowheavy RAM offload |
| Kimi-VL A3B Thinking 2506 | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 96.3 tok/s | Runs welltight fit |
| Ling 3.0 Tiny (8B-A1.3B) | 60Good quality (B) · 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 45.3 tok/s | Runs great | 60Good quality (B) · 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 102 tok/s | Runs great |
| Granite 4.2 30B | 73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 3.9 tok/s | Slowtight fit | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs welloffloads to RAM |
| K2 Horizon 7B | 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 15.5 tok/s | Runs well | 77Great quality (A) · 77 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. | 39.2 tok/s | Runs great |
| Qwen3 32B | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Gemma 3 27B IT | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 tok/s | Slow | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 18.7 tok/s | Runs welloffloads to RAM |
| Devstral Small 1.1 (24B) | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 4.9 tok/s | Slow | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs welltight fit |
| Magistral Small 2507 | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 4.9 tok/s | Slow | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs welltight fit |
| Mellum2.1 Thinking (12B-A2.5B) | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 39.7 tok/s | Runs great | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 94.5 tok/s | Runs great |
| GLM-4 32B 0414 | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs welloffloads to RAM |
| SenseNova Loop 9B Preview | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 14.5 tok/s | Runs well | 75Great quality (A) · 75 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great |
| MedGemma 27B Text IT | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 5.2 tok/s | Slow | 58Good quality (B) · 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 10.1 tok/s | Runs welloffloads to RAM |
| Granite 4.1 30B | 69Good quality (B) · 69 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 4.0 tok/s | Slowtight fit | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 11.8 tok/s | Runs welloffloads to RAM |
| EXAONE 4.0 32B | 69Good quality (B) · 69 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 3.6 tok/s | Slowtight fit | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 11.5 tok/s | Runs welloffloads to RAM |
| Magistral Small 2506 | 65Good quality (B) · 65 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 4.9 tok/s | Slow | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs welltight fit |
| Nemotron 3.5 Bolt E4B | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 31.0 tok/s | Runs great | 70Great quality (A) · 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 48.4 tok/s | Runs great |
| Mistral Small 3.2 24B | 65Good quality (B) · 65 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 tok/s | Slow | 57Good quality (B) · 57 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 12.6 tok/s | Runs welloffloads to RAM |
| Ornith 1.5 9B | 73Great quality (A) · 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 14.5 tok/s | Runs well | 74Great quality (A) · 74 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great |
| Olmo 3.1 32B Instruct | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 11.3 tok/s | Runs welloffloads to RAM |
| Granite 4.0 H Small (32B-A9B) | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 14.7 tok/s | Runs welltight fit | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 39.6 tok/s | Runs welloffloads to RAM |
| Seed-OSS 36B Instruct | 67Good quality (B) · 67 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 tok/s | Slowtight fit | 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 3.3 tok/s | Slowheavy RAM offload |
| Terminal RSI 27B | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 tok/s | Slow | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs welloffloads to RAM |
| Qwen3.5 9B | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 14.5 tok/s | Runs well | 73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great |
| Hermes 4.3 36B | 66Good quality (B) · 66 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 tok/s | Slowtight fit | 75Great quality (A) · 75 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 3.3 tok/s | Slowheavy RAM offload |
| Gemma 4 E4B | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 31.0 tok/s | Runs great | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 48.4 tok/s | Runs great |
| Hemmingway 1.5 | 63Good quality (B) · 63 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 5.2 tok/s | Slow | 56Good quality (B) · 56 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 10.6 tok/s | Runs welloffloads to RAM |
| Mistral Small 3.1 24B | 64Good quality (B) · 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 tok/s | Slow | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 12.6 tok/s | Runs welloffloads to RAM |
| Ornith 1.0 9B | 71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 14.5 tok/s | Runs well | 72Great quality (A) · 72 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great |
| Baichuan M2 32B | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Qwen2.5 32B Instruct | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| LFM2.5 8B-A1B | 52Fair quality (C) · 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 51Fair quality (C) · 51 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 115 tok/s | Runs great |
| Gemma 4 E2B | 61Good quality (B) · 61 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 42.1 tok/s | Runs great | 62Good quality (B) · 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 94.6 tok/s | Runs great |
| Nanbeige4.2 3B | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 33.5 tok/s | Runs great | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 52.2 tok/s | Runs great |
| TMax 9B | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 15.6 tok/s | Runs well | 71Great quality (A) · 71 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 39.5 tok/s | Runs great |
| Mistral Small 3 24B (2501) | 62Good quality (B) · 62 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 4.9 tok/s | Slow | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs welltight fit |
| Nemotron Labs Dory 1.0 4B | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 35.1 tok/s | Runs great | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 54.7 tok/s | Runs great |
| Qwen2.5-VL 32B Instruct | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 4.2 tok/s | Slowtight fit | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 3.7 tok/s | Slowheavy RAM offload |
| InternVL3.5 38B | 63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 4.7 tok/s | Slowtight fit | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 3.0 tok/s | Slowheavy RAM offload |
| Kimi-VL A3B Instruct | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 96.3 tok/s | Runs welltight fit |
| Dolphin Mistral 24B Venice Edition | 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 tok/s | Slow | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 27.5 tok/s | Runs welloffloads to RAM |
| QwQ 32B | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Voxtral Small 24B | 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 tok/s | Slow | 52Fair quality (C) · 52 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs welloffloads to RAM |
| Phi-3.5 MoE (42B-A6.6B) | 47Fair quality (C) · 47 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 25.1 tok/s | Runs welltight fit | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 16.9 tok/s | Slowheavy RAM offload |
| Spark-X2.6 4B | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 34.0 tok/s | Runs great | 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 53.0 tok/s | Runs great |
| Gemma 2 27B IT | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 tok/s | Slow | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 15.1 tok/s | Runs welloffloads to RAM |
| Granite 4.2 8B | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 15.9 tok/s | Runs well | 65Good quality (B) · 65 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great |
| Magistral Small 1.2 (24B) | 69Good quality (B) · 69 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 tok/s | Slow | 60Good quality (B) · 60 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 12.6 tok/s | Runs welloffloads to RAM |
| Dolphin 3.0 R1 Mistral 24B | 58Good quality (B) · 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 4.9 tok/s | Slow | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs welltight fit |
| Olmo 3.1 32B Think | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 11.3 tok/s | Runs welloffloads to RAM |
| OpenReasoning Nemotron 32B | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Skywork OR1 32B | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Falcon-H1 34B Instruct | 60Good quality (B) · 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 4.2 tok/s | Slowtight fit | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 9.4 tok/s | Slowoffloads to RAM |
| MiniCPM5 2B | 59Good quality (B) · 59 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 38.5 tok/s | Runs great | 60Good quality (B) · 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 86.4 tok/s | Runs great |
| Ministral 3 14B | 60Good quality (B) · 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 10.0 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 30.7 tok/s | Runs welltight fit |
| Moonlight 16B-A3B Instruct | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 96.3 tok/s | Runs welltight fit |
| Apriel 1.5 15B Thinker | 60Good quality (B) · 60 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 28.8 tok/s | Runs welltight fit |
| InternVL3 38B | 58Good quality (B) · 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 4.7 tok/s | Slowtight fit | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 3.0 tok/s | Slowheavy RAM offload |
| EXAONE Deep 32B | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 4.4 tok/s | Slowtight fit | 71Great quality (A) · 71 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 3.1 tok/s | Slowheavy RAM offload |
| GLM-Z1 32B (0414) | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs welloffloads to RAM |
| Command R (08-2024) 32B | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 10.2 tok/s | Runs welloffloads to RAM |
| MiMo V2.6 Distill Qwen 9B | 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 14.8 tok/s | Runs well | 75Great quality (A) · 75 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 37.5 tok/s | Runs great |
| DeepSeek Coder V2 Lite (16B) | 50Fair quality (C) · 50 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 41.2 tok/s | Runs great | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 111 tok/s | Runs welltight fit |
| Phi-4 Reasoning Vision 15B | 58Good quality (B) · 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 11.9 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs welltight fit |
| DeepSeek V2 Lite Chat | 38Basic quality (D) · 38 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 41.2 tok/s | Runs great | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 111 tok/s | Runs welltight fit |
| DeepSeek R1 Distill Qwen 32B | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| TMax 4B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 33.2 tok/s | Runs great | 61Good quality (B) · 61 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 51.7 tok/s | Runs great |
| Qwen3.5 4B | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 30.0 tok/s | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 46.7 tok/s | Runs great |
| OLMo 2 32B Instruct | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 10.7 tok/s | Runs welloffloads to RAM |
| Aya Expanse 32B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 10.2 tok/s | Runs welloffloads to RAM |
| Aya Vision 32B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 4.2 tok/s | Slowtight fit | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 9.6 tok/s | Slowoffloads to RAM |
| Llama 3.3 Nemotron Super 49B v1.5 | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 4.6 tok/s | Slowtight fit | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 4.0 tok/s | Slowheavy RAM offload |
| LFM3 1.7B | 56Good quality (B) · 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 41.1 tok/s | Runs great | 56Good quality (B) · 56 of 100: how capable the model is at F16, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 58.0 tok/s | Runs great |
| InternVL3.5 8B | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 16.4 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 41.4 tok/s | Runs great |
| Qwen3-VL 8B Instruct | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 15.9 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 40.3 tok/s | Runs great |
| Devstral Small 2505 | 64Good quality (B) · 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 4.9 tok/s | Slow | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs welltight fit |
| Qwen3 14B | 57Good quality (B) · 57 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 28.9 tok/s | Runs welltight fit |
| Agents-A1 4B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 30.8 tok/s | Runs great | 60Good quality (B) · 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 47.9 tok/s | Runs great |
| OpenHands LM 32B | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Granite 4.0 H Tiny (7B-A1B) | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 56.0 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 119 tok/s | Runs great |
| SenseNova 6.8 Code2Video | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 tok/s | Slow | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 10.1 tok/s | Runs welloffloads to RAM |
| Granite 4.1 8B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 15.9 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great |
| Qwen3 8B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great |
| MiniCPM-V 4.5 8B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 16.1 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 40.6 tok/s | Runs great |
| Gemma 3 12B IT | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs well | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 35.0 tok/s | Runs great |
| Hermes 4 14B | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| Fara 1.5 4B | 55Good quality (B) · 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 30.8 tok/s | Runs great | 58Good quality (B) · 58 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 47.9 tok/s | Runs great |
| Mage-VL 4.7B | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 29.5 tok/s | Runs well | 56Good quality (B) · 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 45.9 tok/s | Runs great |
| InternVL3 14B | 53Fair quality (C) · 53 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 11.9 tok/s | Runs well | 53Fair quality (C) · 53 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs welltight fit |
| Phi-4 (14B) | 53Fair quality (C) · 53 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 12.3 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 29.1 tok/s | Runs welltight fit |
| Qwen3 4B Instruct 2507 | 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 34.7 tok/s | Runs great | 54Fair quality (C) · 54 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 54.1 tok/s | Runs great |
| GLM-4.1V 9B Thinking | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 13.6 tok/s | Runs well | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 41.6 tok/s | Runs great |
| Qwen2.5 Coder 32B Instruct | 63Good quality (B) · 63 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 63Good quality (B) · 63 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| MiMo-VL 7B RL | 55Good quality (B) · 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 16.8 tok/s | Runs well | 56Good quality (B) · 56 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 42.5 tok/s | Runs great |
| LFM2 8B-A1B | 38Basic quality (D) · 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 38Basic quality (D) · 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 93.5 tok/s | Runs great |
| Nemotron Nano 9B v2 | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 15.7 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 39.7 tok/s | Runs great |
| Qwen3-VL 4B Instruct | 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 31.5 tok/s | Runs great | 54Fair quality (C) · 54 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 49.0 tok/s | Runs great |
| QwQ 32B Preview | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Qwen2.5 14B Instruct | 52Fair quality (C) · 52 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 12.3 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 29.1 tok/s | Runs welltight fit |
| Granite 4.2 3B | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 38.2 tok/s | Runs great | 52Fair quality (C) · 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 59.5 tok/s | Runs great |
| Ministral 3 14B Reasoning | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 10.0 tok/s | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 30.7 tok/s | Runs welltight fit |
| UI-Mate 9B | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 14.8 tok/s | Runs well | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 37.5 tok/s | Runs great |
| UI-Venus 2 9B | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 14.8 tok/s | Runs well | 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 37.5 tok/s | Runs great |
| MiMo 7B RL | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 17.8 tok/s | Runs well | 55Good quality (B) · 55 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 37.9 tok/s | Runs great |
| Ministral 3 8B | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 15.7 tok/s | Runs well | 54Fair quality (C) · 54 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 39.6 tok/s | Runs great |
| Qwen2.5 14B Instruct 1M | 50Fair quality (C) · 50 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| LFM3-VL 2B | 49Fair quality (C) · 49 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 50.5 tok/s | Runs great | 50Fair quality (C) · 50 of 100: how capable the model is at F16, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 52.3 tok/s | Runs great |
| Qwen3 4B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 34.7 tok/s | Runs great | 50Fair quality (C) · 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 54.1 tok/s | Runs great |
| InternVL3 8B | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 17.6 tok/s | Runs well | 55Good quality (B) · 55 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 37.4 tok/s | Runs great |
| Phi-4 Reasoning Plus (14B) | 60Good quality (B) · 60 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 12.3 tok/s | Runs well | 68Good quality (B) · 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 29.2 tok/s | Runs welltight fit |
| Llama 3.1 Nemotron Nano 8B v1 | 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 52Fair quality (C) · 52 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Gemma 3n E4B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 34.9 tok/s | Runs great | 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 54.4 tok/s | Runs great |
| Mistral Nemo 12B Instruct | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs well | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 35.0 tok/s | Runs great |
| Yi 1.5 34B Chat | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 tok/s | Slowtight fit | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 9.2 tok/s | Slowoffloads to RAM |
| Nous Hermes 2 Mixtral 8x7B DPO | 36Basic quality (D) · 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 16.8 tok/s | Runs welltight fit | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 8.5 tok/s | Slowheavy RAM offload |
| Nemotron 3 Nano 4B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 35.2 tok/s | Runs great | 50Fair quality (C) · 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 54.8 tok/s | Runs great |
| Falcon 3 10B Instruct | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 13.5 tok/s | Runs well | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 41.5 tok/s | Runs great |
| Phi-4 Reasoning (14B) | 58Good quality (B) · 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 12.3 tok/s | Runs well | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 29.2 tok/s | Runs welltight fit |
| MiniCPM4 8B | 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 52Fair quality (C) · 52 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great |
| Jamba 1.5 Mini | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 18.0 tok/s | Runs welltight fit | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 9.1 tok/s | Slowheavy RAM offload |
| Hunyuan 7B Instruct | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 18.6 tok/s | Runs well | 51Fair quality (C) · 51 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 39.6 tok/s | Runs great |
| Pixtral 12B (2409) | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 11.3 tok/s | Runs well | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 34.5 tok/s | Runs great |
| Mage-VL-Cowork 4.7B | 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 29.5 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 45.9 tok/s | Runs great |
| Mixtral 8x7B Instruct | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 16.8 tok/s | Runs welltight fit | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 8.5 tok/s | Slowheavy RAM offload |
| Granite 4.1 3B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 41.1 tok/s | Runs great | 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 64.0 tok/s | Runs great |
| North Micro Vision | 43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 39.1 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 40.5 tok/s | Runs great |
| Gemma 2 9B IT | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 15.1 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 38.2 tok/s | Runs great |
| Olmo 3 7B Instruct | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs well | 49Fair quality (C) · 49 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 40.6 tok/s | Runs great |
| Qwen2.5 7B Instruct | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 51Fair quality (C) · 51 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| Qwen2.5-VL 7B Instruct | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 16.8 tok/s | Runs well | 50Fair quality (C) · 50 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 42.6 tok/s | Runs great |
| MiniCPM-o 2.6 8B | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 16.1 tok/s | Runs well | 50Fair quality (C) · 50 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 40.7 tok/s | Runs great |
| DeepSeek R1 0528 Qwen3 8B | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great |
| Qwen2 57B-A14B Instruct | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 15.6 tok/s | Runs welltight fit | 42Fair quality (C) · 42 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 10.4 tok/s | Slowheavy RAM offload |
| GLM-4 9B 0414 | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 14.9 tok/s | Runs well | 50Fair quality (C) · 50 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 37.6 tok/s | Runs great |
| OLMo 2 13B Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 10.2 tok/s | Runs well | 36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 40.3 tok/s | Runs welltight fit |
| NorthCUA Nano | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 14.8 tok/s | Runs well | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 45.4 tok/s | Runs great |
| Foundation-Sec 1.1 8B Instruct | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| InternVL2.5 8B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 43.7 tok/s | Runs great |
| Phi-3.5 Mini Instruct | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 57.0 tok/s | Runs great |
| Qwen3 4B Thinking 2507 | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 34.7 tok/s | Runs great | 56Good quality (B) · 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 54.1 tok/s | Runs great |
| Phi-3 Medium 128K Instruct | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 10.0 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 39.6 tok/s | Runs great |
| Phi-4 Mini (3.8B) | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 36.4 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 56.7 tok/s | Runs great |
| Unlimited OCR 2 (3B-A0.6B) | 44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 44.1 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 96.6 tok/s | Runs great |
| Apriel Nemotron 15B Thinker | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 12.1 tok/s | Runs well | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 36.9 tok/s | Runs welltight fit |
| Qwen2.5 7B Instruct 1M | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 49Fair quality (C) · 49 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| AceReason Nemotron 14B | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| OpenReasoning Nemotron 14B | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| Qwen2.5 Coder 14B Instruct | 55Good quality (B) · 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| LFM2.5 2.6B | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 80.6 tok/s | Runs great |
| Falcon-H1 7B Instruct | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 18.4 tok/s | Runs well | 49Fair quality (C) · 49 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 39.1 tok/s | Runs great |
| Ministral 3 3B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 36.3 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 56.5 tok/s | Runs great |
| InternLM3 8B Instruct | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 15.9 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 40.1 tok/s | Runs great |
| GLiDE 9B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 17.6 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.5 tok/s | Runs great |
| DeepSeek MoE 16B Chat | 24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 tok/s | Runs great | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 120 tok/s | Runs great |
| Hermes 3 (Llama 3.1 8B) | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 46Fair quality (C) · 46 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Llama 3.1 8B Instruct | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 46Fair quality (C) · 46 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| LFM2.5-VL 3B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 44.8 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 69.8 tok/s | Runs great |
| Jamba Reasoning 3B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 43.6 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 68.0 tok/s | Runs great |
| Solar Pro Preview 22B | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 10.4 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 31.7 tok/s | Runs welltight fit |
| Phi-3 Medium (14B) | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 10.0 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 39.6 tok/s | Runs great |
| Llama 3.1 Nemotron Nano 4B v1.1 | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 31.0 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 48.3 tok/s | Runs great |
| Qwen1.5 32B Chat | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 tok/s | Slowtight fit | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. | 3.9 tok/s | Slowheavy RAM offload |
| GLM-4 9B Chat 1M | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 14.7 tok/s | Runs well | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 45.1 tok/s | Runs great |
| Gemma 3 4B IT | 42Fair quality (C) · 42 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 32.5 tok/s | Runs great | 45Fair quality (C) · 45 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 50.6 tok/s | Runs great |
| DeepCoder 14B Preview | 53Fair quality (C) · 53 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| LLaVA 1.6 34B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 4.0 tok/s | Slowtight fit | 30Basic quality (D) · 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 9.0 tok/s | Slowoffloads to RAM |
| SmolLM3 3B | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 45.3 tok/s | Runs great | 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 70.7 tok/s | Runs great |
| MiniCPM-V 2.6 8B | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.2 tok/s | Runs well | 46Fair quality (C) · 46 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 43.6 tok/s | Runs great |
| Gemma 3n E2B | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 48.5 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 109 tok/s | Runs great |
| Phi-3.5 Vision Instruct | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 33.7 tok/s | Runs great | 38Basic quality (D) · 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 52.5 tok/s | Runs great |
| Granite 3.2 8B Instruct | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 43.2 tok/s | Runs great |
| Granite 3.3 8B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 43.2 tok/s | Runs great |
| GLM-4 9B Chat (2024) | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 14.9 tok/s | Runs well | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 45.5 tok/s | Runs great |
| Ministral 8B (2410) | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Hunyuan 4B Instruct | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 33.1 tok/s | Runs great | 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 51.6 tok/s | Runs great |
| Seed-Coder 8B Reasoning | 55Good quality (B) · 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 16.9 tok/s | Runs well | 56Good quality (B) · 56 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 42.8 tok/s | Runs great |
| Jamba2 3B | 40Fair quality (C) · 40 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 38.1 tok/s | Runs great | 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 71.8 tok/s | Runs great |
| MedGemma 4B IT | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 32.5 tok/s | Runs great | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 50.6 tok/s | Runs great |
| EXAONE 3.5 7.8B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.9 tok/s | Runs well | 45Fair quality (C) · 45 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 37.9 tok/s | Runs great |
| DeepHermes 3 Llama 3 8B Preview | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Dolphin 3.0 Llama 3.1 8B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Tülu 3 8B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Apertus v1.5 8B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 15.7 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 39.7 tok/s | Runs great |
| Falcon 3 7B Instruct | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 18.7 tok/s | Runs well | 45Fair quality (C) · 45 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 39.8 tok/s | Runs great |
| Phi-4 Multimodal Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 25.1 tok/s | Runs well | 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 39.1 tok/s | Runs great |
| Phi-3 Mini 128K Instruct | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 57.0 tok/s | Runs great |
| Qwen3.5 2B | 39Basic quality (D) · 39 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 44.2 tok/s | Runs great |
| MiniCPM-V 4.7 1B | 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 45.9 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 64.7 tok/s | Runs great |
| Granite 4.0 Micro (3B) | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 41.1 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 64.0 tok/s | Runs great |
| DeepSeek R1 Distill Qwen 14B | 52Fair quality (C) · 52 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| Qwen2.5-Omni 7B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 13.0 tok/s | Runs well | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 39.9 tok/s | Runs great |
| Qwen1.5 MoE A2.7B Chat | 22Basic quality (D) · 22 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 44.1 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 103 tok/s | Runs welltight fit |
| Unlimited OCR (3B-A0.6B) | 40Fair quality (C) · 40 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 44.1 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 96.6 tok/s | Runs great |
| GLM-Z1 9B (0414) | 55Good quality (B) · 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 14.9 tok/s | Runs well | 56Good quality (B) · 56 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 37.6 tok/s | Runs great |
| JanusCoder 14B | 52Fair quality (C) · 52 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 12.2 tok/s | Runs well | 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 29.0 tok/s | Runs welltight fit |
| TMax 2B | 39Basic quality (D) · 39 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 51.5 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 53.4 tok/s | Runs great |
| Granite 3.1 8B Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 43.2 tok/s | Runs great |
| Mistral Small 22B (2409) | 37Basic quality (D) · 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 10.3 tok/s | Runs well | 37Basic quality (D) · 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 31.6 tok/s | Runs welltight fit |
| EXAONE Deep 7.8B | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 17.9 tok/s | Runs well | 55Good quality (B) · 55 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 37.9 tok/s | Runs great |
| Seed-Coder 8B Instruct | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 16.9 tok/s | Runs well | 54Fair quality (C) · 54 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 42.8 tok/s | Runs great |
| Phi-3 Small 8K Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 18.9 tok/s | Runs well | 43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 40.1 tok/s | Runs great |
| Dolphin 2.9.4 Llama 3.1 8B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Hermes 2 Pro Llama 3 8B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| OpenChat 3.6 8B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| InternVL2 8B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 43.7 tok/s | Runs great |
| Olmo 3 7B Think | 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs well | 53Fair quality (C) · 53 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 40.6 tok/s | Runs great |
| OLMoE 1B-7B Instruct | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 45.3 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 102 tok/s | Runs great |
| Falcon-H1R 7B | 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 18.4 tok/s | Runs well | 55Good quality (B) · 55 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 39.1 tok/s | Runs great |
| Yi 34B Chat | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 tok/s | Slowtight fit | 28Basic quality (D) · 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 9.2 tok/s | Slowoffloads to RAM |
| Qwen2 7B Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| Llama 3.2 11B Vision Instruct | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 13.1 tok/s | Runs well | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 40.1 tok/s | Runs great |
| Qwen2-VL 7B Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 16.8 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 42.6 tok/s | Runs great |
| Qwen3-VL 2B Instruct | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 45.5 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 47.1 tok/s | Runs great |
| Phi-3 Vision 128K Instruct | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 33.7 tok/s | Runs great | 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 52.5 tok/s | Runs great |
| Command R7B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Qwen1.5 14B Chat | 28Basic quality (D) · 28 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 12.7 tok/s | Runs well | 23Basic quality (D) · 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 22.8 tok/s | Runs welloffloads to RAM |
| Mistral-NeMo-Minitron 8B Instruct | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 16.6 tok/s | Runs well | 40Fair quality (C) · 40 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 42.0 tok/s | Runs great |
| Aya Vision 8B | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 16.2 tok/s | Runs well | 42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 40.9 tok/s | Runs great |
| Phi-3 Mini 4K Instruct | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 36.6 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 57.0 tok/s | Runs great |
| Qwen2.5 Coder 7B Instruct | 52Fair quality (C) · 52 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 54Fair quality (C) · 54 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| AFM 4.5B | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 30.2 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 47.1 tok/s | Runs great |
| Llama 3 8B Instruct | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 40Fair quality (C) · 40 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Granite 4.0 H Micro | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 43.8 tok/s | Runs great | 38Basic quality (D) · 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 68.2 tok/s | Runs great |
| Qianfan-OCR 4B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 29.5 tok/s | Runs well | 50Fair quality (C) · 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 45.9 tok/s | Runs great |
| OpenReasoning Nemotron 7B | 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 53Fair quality (C) · 53 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| MiniCPM3 4B | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 34.3 tok/s | Runs great | 38Basic quality (D) · 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 53.5 tok/s | Runs great |
| Llama 3.2 3B Instruct | 33Basic quality (D) · 33 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 43.5 tok/s | Runs great | 35Basic quality (D) · 35 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 67.8 tok/s | Runs great |
| Command R7B Arabic | 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 40Fair quality (C) · 40 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| MiniCPM5 1B | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 65.8 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 92.9 tok/s | Runs great |
| Foundation-Sec 8B Reasoning | 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 50Fair quality (C) · 50 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Falcon-H1 3B Instruct | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 44.3 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 69.1 tok/s | Runs great |
| InternLM2.5 7B Chat | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 18.0 tok/s | Runs well | 39Basic quality (D) · 39 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 38.3 tok/s | Runs great |
| Aya Expanse 8B | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Apertus 8B Instruct | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 43.9 tok/s | Runs great |
| InternLM2.5 20B Chat | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 11.5 tok/s | Runs well | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 35.4 tok/s | Runs welltight fit |
| Nous Hermes 2 SOLAR 10.7B | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 13.0 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 39.9 tok/s | Runs great |
| SOLAR 10.7B Instruct | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 13.0 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 39.9 tok/s | Runs great |
| Reka Flash 3.1 (21B) | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 11.0 tok/s | Runs well | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 33.6 tok/s | Runs welltight fit |
| Phi-4 Mini Reasoning (3.8B) | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 36.4 tok/s | Runs great | 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 56.7 tok/s | Runs great |
| Granite 3.0 8B Instruct | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 43.2 tok/s | Runs great |
| Voxtral Mini 3B | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs well | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 46.5 tok/s | Runs great |
| Codestral 22B | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 10.3 tok/s | Runs well | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 31.6 tok/s | Runs welltight fit |
| OLMo 2 7B Instruct | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 19.1 tok/s | Runs well | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 40.6 tok/s | Runs great |
| LFM2 2.6B | 32Basic quality (D) · 32 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 44.9 tok/s | Runs great | 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 84.7 tok/s | Runs great |
| S5 nano | 35Basic quality (D) · 35 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 30.0 tok/s | Runs well | 37Basic quality (D) · 37 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 46.7 tok/s | Runs great |
| DeepSeek R1 Distill Llama 8B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Mistral 7B Instruct v0.3 | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 37Basic quality (D) · 37 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 40.9 tok/s | Runs great |
| Qwen2.5 3B Instruct | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 45.2 tok/s | Runs great | 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 70.4 tok/s | Runs great |
| Dolphin 2.9 Llama 3 8B | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 36Basic quality (D) · 36 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| LightOnOCR-3 4B | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 30.8 tok/s | Runs great | 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 47.9 tok/s | Runs great |
| Qwen2.5-VL 3B Instruct | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 37.2 tok/s | Runs great | 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 58.0 tok/s | Runs great |
| Hermes 3 Llama 3.2 3B | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 43.5 tok/s | Runs great | 32Basic quality (D) · 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 67.8 tok/s | Runs great |
| Zamba2 7B Instruct | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 19.0 tok/s | Runs well | 37Basic quality (D) · 37 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 40.3 tok/s | Runs great |
| Fara 7B | 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 16.8 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 42.6 tok/s | Runs great |
| EXAONE 3.0 7.8B Instruct | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 17.9 tok/s | Runs well | 36Basic quality (D) · 36 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 45.2 tok/s | Runs great |
| Chandra OCR 2 | 45Fair quality (C) · 45 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 26.4 tok/s | Runs well | 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 41.1 tok/s | Runs great |
| CogVLM2 Llama 3 Chat 19B | 26Basic quality (D) · 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 11.8 tok/s | Runs well | 26Basic quality (D) · 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 21.2 tok/s | Runs welloffloads to RAM |
| EXAONE 3.5 2.4B | 29Basic quality (D) · 29 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 40.2 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 41.6 tok/s | Runs great |
| Qwen3 1.7B | 29Basic quality (D) · 29 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 47.7 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at F16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 49.4 tok/s | Runs great |
| WeVisOne-VL 4B | 42Fair quality (C) · 42 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 28.9 tok/s | Runs well | 45Fair quality (C) · 45 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 45.1 tok/s | Runs great |
| Hermes 2 Pro Mistral 7B | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| Starling LM 7B Beta | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| LFM2.5 1.2B Instruct | 32Basic quality (D) · 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 60.8 tok/s | Runs great | 32Basic quality (D) · 32 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 85.7 tok/s | Runs great |
| Falcon3 Mamba 7B Instruct | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 19.2 tok/s | Runs well | 37Basic quality (D) · 37 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 40.8 tok/s | Runs great |
| LLM4Decompile 22B V2 | 40Fair quality (C) · 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 10.3 tok/s | Runs well | 40Fair quality (C) · 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 31.6 tok/s | Runs welltight fit |
| DeepSeek Coder 33B Instruct | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 4.2 tok/s | Slowtight fit | 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Granite 3.2 2B Instruct | 29Basic quality (D) · 29 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 45.6 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 86.0 tok/s | Runs great |
| Granite 3.3 2B | 29Basic quality (D) · 29 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 45.6 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 86.0 tok/s | Runs great |
| OpenChat 3.5 (0106) | 33Basic quality (D) · 33 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 34Basic quality (D) · 34 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| Falcon 3 3B Instruct | 28Basic quality (D) · 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 43.2 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 67.4 tok/s | Runs great |
| MiniCPM-OCR 1.0 1.3B | 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 54.7 tok/s | Runs great | 42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 77.2 tok/s | Runs great |
| TeleOCR | 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 50.1 tok/s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 70.6 tok/s | Runs great |
| EXAONE 4.0 1.2B | 30Basic quality (D) · 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 55.5 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 78.4 tok/s | Runs great |
| DeepSeek Prover V2 7B | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 20.2 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 51.1 tok/s | Runs great |
| PaddleOCR-VL 1.6 0.9B | 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 74.1 tok/s | Runs great | 42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 105 tok/s | Runs great |
| Yi 1.5 9B Chat | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 15.8 tok/s | Runs well | 34Basic quality (D) · 34 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 40.0 tok/s | Runs great |
| Mistral 7B Instruct v0.2 | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 33Basic quality (D) · 33 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| Nous Hermes 2 Mistral 7B DPO | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 33Basic quality (D) · 33 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| OpenChat 3.5 (1210) | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 33Basic quality (D) · 33 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| InternLM2 Chat 20B | 29Basic quality (D) · 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 11.5 tok/s | Runs well | 29Basic quality (D) · 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 35.4 tok/s | Runs welltight fit |
| Aya 23 8B | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 17.4 tok/s | Runs well | 32Basic quality (D) · 32 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Orca 2 13B | 26Basic quality (D) · 26 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 10.7 tok/s | Runs well | 23Basic quality (D) · 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 42.5 tok/s | Runs welltight fit |
| WeVisOne VL 2B | 26Basic quality (D) · 26 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 39.7 tok/s | Runs great | 27Basic quality (D) · 27 of 100: how capable the model is at F16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 41.1 tok/s | Runs great |
| Gemma 1.1 7B IT | 26Basic quality (D) · 26 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 16.4 tok/s | Runs well | 26Basic quality (D) · 26 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 50.1 tok/s | Runs great |
| Granite 3.1 2B Instruct | 27Basic quality (D) · 27 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 45.6 tok/s | Runs great | 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 86.0 tok/s | Runs great |
| DeepSeek R1 Distill Qwen 7B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 45Fair quality (C) · 45 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| Gemma 2 2B | 27Basic quality (D) · 27 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 44.2 tok/s | Runs great | 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 83.4 tok/s | Runs great |
| CodeGeeX4 All 9B | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 14.9 tok/s | Runs well | 43Fair quality (C) · 43 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 45.5 tok/s | Runs great |
| Baichuan 2 13B Chat | 24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 10.0 tok/s | Runs well | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 39.8 tok/s | Runs welltight fit |
| Granite 4.0 H 1B | 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 48.7 tok/s | Runs great | 28Basic quality (D) · 28 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 68.7 tok/s | Runs great |
| OpenHermes 2.5 Mistral 7B | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 31Basic quality (D) · 31 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| Starling LM 7B Alpha | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 31Basic quality (D) · 31 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| RecurrentGemma 9B IT | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 14.5 tok/s | Runs well | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 44.4 tok/s | Runs great |
| Llama 2 13B Chat | 24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 10.7 tok/s | Runs well | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 42.5 tok/s | Runs welltight fit |
| SmolVLM2 2.2B Instruct | 23Basic quality (D) · 23 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great | 24Basic quality (D) · 24 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 44.6 tok/s | Runs great |
| Granite Code 34B Instruct | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 tok/s | Slowtight fit | 28Basic quality (D) · 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. | 10.6 tok/s | Runs welloffloads to RAM |
| Qwen1.5 7B Chat | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 23.4 tok/s | Runs well | 25Basic quality (D) · 25 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 45.8 tok/s | Runs welltight fit |
| Falcon-H1 1.5B Instruct | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 45.9 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 64.7 tok/s | Runs great |
| Granite Vision 3.2 2B | 24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 46.9 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 73.0 tok/s | Runs great |
| Granite 3.0 2B Instruct | 25Basic quality (D) · 25 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 43.9 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 82.8 tok/s | Runs great |
| Code Llama 13B Instruct | 28Basic quality (D) · 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 10.7 tok/s | Runs well | 21Basic quality (D) · 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. | 28.3 tok/s | Runs welloffloads to RAM |
| Mistral 7B Instruct v0.1 | 28Basic quality (D) · 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 29Basic quality (D) · 29 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| LFM2 1.2B | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 60.8 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 85.7 tok/s | Runs great |
| SmolVLM 2.2B Instruct | 21Basic quality (D) · 21 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 44.6 tok/s | Runs great |
| DeepSeek Coder 6.7B Instruct | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 20.7 tok/s | Runs well | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 63.4 tok/s | Runs great |
| Gemma 7B IT | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 21.1 tok/s | Runs well | 24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 50.1 tok/s | Runs great |
| Code Llama 34B Instruct | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 tok/s | Slowtight fit | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 39/100 (quality, fit, speed and memory use) sets the ranking. | 3.8 tok/s | Slowheavy RAM offload |
| Qwen3.5 0.8B | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 81.7 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 115 tok/s | Runs great |
| Granite 3.1 3B-A800M Instruct | 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 66.4 tok/s | Runs great | 20Basic quality (D) · 20 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 81.0 tok/s | Runs great |
| Qwen2.5-Omni 3B | 25Basic quality (D) · 25 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 32.6 tok/s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 39.3 tok/s | Runs great |
| Qwen2-VL 2B Instruct | 23Basic quality (D) · 23 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 43.8 tok/s | Runs great | 24Basic quality (D) · 24 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 45.4 tok/s | Runs great |
| Qwen2.5 1.5B Instruct | 24Basic quality (D) · 24 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 46.2 tok/s | Runs great | 24Basic quality (D) · 24 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 65.1 tok/s | Runs great |
| LFM2-VL 1.6B | 24Basic quality (D) · 24 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 45.0 tok/s | Runs great | 24Basic quality (D) · 24 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 63.5 tok/s | Runs great |
| Qwen2.5 Coder 3B Instruct | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 45.2 tok/s | Runs great | 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 70.4 tok/s | Runs great |
| StableLM 2 12B Chat | 26Basic quality (D) · 26 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 11.5 tok/s | Runs well | 26Basic quality (D) · 26 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 35.2 tok/s | Runs great |
| LLaVA 1.5 13B | 20Basic quality (D) · 20 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 10.5 tok/s | Runs well | 18Basic quality (D) · 18 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 41.4 tok/s | Runs welltight fit |
| Falcon Mamba 7B Instruct | 28Basic quality (D) · 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 19.2 tok/s | Runs well | 29Basic quality (D) · 29 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 40.8 tok/s | Runs great |
| WizardLM 13B v1.2 | 20Basic quality (D) · 20 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 10.7 tok/s | Runs well | 18Basic quality (D) · 18 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 42.5 tok/s | Runs welltight fit |
| CodeGemma 7B | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 16.4 tok/s | Runs well | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 50.1 tok/s | Runs great |
| SmolLM2 1.7B Instruct | 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 41.6 tok/s | Runs great | 20Basic quality (D) · 20 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 58.7 tok/s | Runs great |
| Hunyuan 1.8B Instruct | 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 39.7 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 56.0 tok/s | Runs great |
| Yi Coder 9B Chat | 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 15.8 tok/s | Runs well | 38Basic quality (D) · 38 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 40.0 tok/s | Runs great |
| LLM4Decompile 6.7B V2 | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 20.7 tok/s | Runs well | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 63.4 tok/s | Runs great |
| Zamba2 2.7B Instruct | 21Basic quality (D) · 21 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 43.4 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 81.8 tok/s | Runs great |
| Neural Chat 7B v3.3 | 23Basic quality (D) · 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 24.9 tok/s | Runs well | 27Basic quality (D) · 27 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| Granite 3.0 3B-A800M Instruct | 18Basic quality (D) · 18 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 66.4 tok/s | Runs great | 18Basic quality (D) · 18 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 81.0 tok/s | Runs great |
| Yi 1.5 6B Chat | 23Basic quality (D) · 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 29.8 tok/s | Runs well | 27Basic quality (D) · 27 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 49.0 tok/s | Runs great |
| LLaVA 1.6 Mistral 7B | 23Basic quality (D) · 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 23.9 tok/s | Runs well | 27Basic quality (D) · 27 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 39.2 tok/s | Runs great |
| Llama 3.2 1B Instruct | 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 57.3 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 80.9 tok/s | Runs great |
| LLM4Decompile 9B V2 | 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 15.8 tok/s | Runs well | 36Basic quality (D) · 36 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 40.0 tok/s | Runs great |
| SQLCoder 7B v2 | 28Basic quality (D) · 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 20.7 tok/s | Runs well | 28Basic quality (D) · 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 63.4 tok/s | Runs great |
| Zephyr 7B Beta | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 24.9 tok/s | Runs well | 25Basic quality (D) · 25 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| Nemotron Mini 4B Instruct | 18Basic quality (D) · 18 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 52.0 tok/s | Runs great |
| StarCoder2 15B Instruct | 33Basic quality (D) · 33 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 11.3 tok/s | Runs well | 33Basic quality (D) · 33 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 34.6 tok/s | Runs great |
| InternLM2.5 1.8B Chat | 17Basic quality (D) · 17 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 51.3 tok/s | Runs great | 18Basic quality (D) · 18 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 53.1 tok/s | Runs great |
| Orca 2 7B | 15Basic quality (D) · 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 34.0 tok/s | Runs great | 21Basic quality (D) · 21 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Gemma 3 1B | 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 71.1 tok/s | Runs great | 20Basic quality (D) · 20 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 100 tok/s | Runs great |
| Vicuna 13B v1.5 | 16Basic quality (D) · 16 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 13.9 tok/s | Runs well | 16Basic quality (D) · 16 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 42.5 tok/s | Runs welltight fit |
| StableLM Zephyr 3B | 15Basic quality (D) · 15 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 41.2 tok/s | Runs great | 16Basic quality (D) · 16 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 77.7 tok/s | Runs great |
| Qwen2.5-Math 7B Instruct | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 18.3 tok/s | Runs well | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 38.9 tok/s | Runs great |
| Mathstral 7B | 32Basic quality (D) · 32 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 19.3 tok/s | Runs well | 33Basic quality (D) · 33 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 40.9 tok/s | Runs great |
| Codestral Mamba 7B | 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 19.2 tok/s | Runs well | 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 40.7 tok/s | Runs great |
| LFM2 700M | 18Basic quality (D) · 18 of 100: how capable the model is at F16, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 44.3 tok/s | Runs great | 18Basic quality (D) · 18 of 100: how capable the model is at F16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 136 tok/s | Runs great |
| Llama 2 7B Chat | 14Basic quality (D) · 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 34.0 tok/s | Runs great | 19Basic quality (D) · 19 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Phi-2 2.7B | 15Basic quality (D) · 15 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 41.5 tok/s | Runs great | 16Basic quality (D) · 16 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 78.3 tok/s | Runs great |
| Falcon 3 1B Instruct | 16Basic quality (D) · 16 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 42.6 tok/s | Runs great | 16Basic quality (D) · 16 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 60.1 tok/s | Runs great |
| InternLM2 Math Plus 20B | 28Basic quality (D) · 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 11.5 tok/s | Runs well | 28Basic quality (D) · 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 35.4 tok/s | Runs welltight fit |
| GLiDE 2B | 16Basic quality (D) · 16 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 51.5 tok/s | Runs great | 17Basic quality (D) · 17 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 53.4 tok/s | Runs great |
| Granite Code 8B Instruct (128K) | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 tok/s | Runs well | 31Basic quality (D) · 31 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 43.9 tok/s | Runs great |
| Baichuan 2 7B Chat | 14Basic quality (D) · 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 30.5 tok/s | Runs great | 19Basic quality (D) · 19 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 39.5 tok/s | Runs great |
| Spark X2.6 1.7B | 16Basic quality (D) · 16 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 41.6 tok/s | Runs great | 16Basic quality (D) · 16 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 58.7 tok/s | Runs great |
| InternLM2 Math Plus 7B | 30Basic quality (D) · 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 18.0 tok/s | Runs well | 31Basic quality (D) · 31 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 45.6 tok/s | Runs great |
| Qwen2 1.5B Instruct | 16Basic quality (D) · 16 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 46.2 tok/s | Runs great | 16Basic quality (D) · 16 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 65.1 tok/s | Runs great |
| DeepScaleR 1.5B Preview | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 39.9 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 56.4 tok/s | Runs great |
| DeepSeek LLM 7B Chat | 12Basic quality (D) · 12 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 33.2 tok/s | Runs great | 17Basic quality (D) · 17 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 42.9 tok/s | Runs great |
| Qwen3 0.6B | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 43.7 tok/s | Runs great | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 134 tok/s | Runs great |
| Qwen2.5 Coder 1.5B Instruct | 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 46.2 tok/s | Runs great | 26Basic quality (D) · 26 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 65.1 tok/s | Runs great |
| Yi 6B Chat | 14Basic quality (D) · 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 37.9 tok/s | Runs great | 19Basic quality (D) · 19 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 49.0 tok/s | Runs great |
| SmolLM 1.7B Instruct | 14Basic quality (D) · 14 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 41.6 tok/s | Runs great | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 58.7 tok/s | Runs great |
| Falcon 40B Instruct | 11Basic quality (D) · 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 5.5 tok/s | Slow | 11Basic quality (D) · 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 35/100 (quality, fit, speed and memory use) sets the ranking. | 6.6 tok/s | Slowheavy RAM offload |
| OLMo 2 1B Instruct | 14Basic quality (D) · 14 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 48.0 tok/s | Runs great | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 67.8 tok/s | Runs great |
| LLaVA 1.5 7B | 11Basic quality (D) · 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 32.5 tok/s | Runs great | 15Basic quality (D) · 15 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 42.0 tok/s | Runs great |
| Stable Code Instruct 3B | 19Basic quality (D) · 19 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 41.2 tok/s | Runs great | 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 77.7 tok/s | Runs great |
| Granite Code 3B Instruct (128K) | 19Basic quality (D) · 19 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 40.1 tok/s | Runs great | 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 62.6 tok/s | Runs great |
| Code Llama 7B Instruct | 16Basic quality (D) · 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 34.0 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 63.4 tok/s | Runs great |
| StarCoder2 15B (base) | 26Basic quality (D) · 26 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 11.3 tok/s | Runs well | 26Basic quality (D) · 26 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 34.6 tok/s | Runs great |
| Vicuna 7B v1.5 | 9Basic quality (D) · 9 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 34.0 tok/s | Runs great | 13Basic quality (D) · 13 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 44.0 tok/s | Runs great |
| Falcon-H1 0.5B Instruct | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 63.0 tok/s | Runs great | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 193 tok/s | Runs great |
| LFM2.5 350M | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 93.6 tok/s | Runs great | 14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 287 tok/s | Runs great |
| Gemma 2B IT | 11Basic quality (D) · 11 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 46.0 tok/s | Runs great | 12Basic quality (D) · 12 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 86.7 tok/s | Runs great |
| DeepSeek R1 Distill Qwen 1.5B | 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 39.9 tok/s | Runs great | 22Basic quality (D) · 22 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 56.4 tok/s | Runs great |
| StableLM 2 1.6B Chat | 10Basic quality (D) · 10 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 43.4 tok/s | Runs great | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 61.2 tok/s | Runs great |
| StableLM 2 Zephyr 1.6B | 10Basic quality (D) · 10 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 43.4 tok/s | Runs great | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 61.2 tok/s | Runs great |
| Granite Code 20B Instruct | 25Basic quality (D) · 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs well | 25Basic quality (D) · 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 35.0 tok/s | Runs great |
| OLMo 7B Instruct | 9Basic quality (D) · 9 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 33.3 tok/s | Runs great | 13Basic quality (D) · 13 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 43.1 tok/s | Runs great |
| GLiDE 0.8B | 11Basic quality (D) · 11 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 43.7 tok/s | Runs great | 11Basic quality (D) · 11 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 134 tok/s | Runs great |
| LFM2-VL 450M | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 72.8 tok/s | Runs great | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 223 tok/s | Runs great |
| LFM2 350M | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 93.6 tok/s | Runs great | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 287 tok/s | Runs great |
| Granite 4.0 H 350M | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 96.4 tok/s | Runs great | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 295 tok/s | Runs great |
| d1 2B | 20Basic quality (D) · 20 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 50.5 tok/s | Runs great | 21Basic quality (D) · 21 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 52.3 tok/s | Runs great |
| WizardMath 7B v1.1 | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 24.9 tok/s | Runs well | 25Basic quality (D) · 25 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 41.0 tok/s | Runs great |
| SmolVLM 500M Instruct | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 64.3 tok/s | Runs great | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 197 tok/s | Runs great |
| InternLM 7B | 8Basic quality (D) · 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 31.3 tok/s | Runs great | 11Basic quality (D) · 11 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 40.5 tok/s | Runs great |
| Qwen2.5-Math 1.5B Instruct | 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 46.2 tok/s | Runs great | 20Basic quality (D) · 20 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 65.1 tok/s | Runs great |
| Qwen2.5 0.5B Instruct | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 66.9 tok/s | Runs great | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 205 tok/s | Runs great |
| Hunyuan 0.5B Instruct | 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 60.7 tok/s | Runs great | 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 186 tok/s | Runs great |
| Command R 35B (v01) | 37Basic quality (D) · 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 5.6 tok/s | Slowover memory limit | 37Basic quality (D) · 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Yi Coder 1.5B Chat | 16Basic quality (D) · 16 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 48.0 tok/s | Runs great | 16Basic quality (D) · 16 of 100: how capable the model is at F16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 67.8 tok/s | Runs great |
| LFM2.5 230M | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 142 tok/s | Runs great | 10Basic quality (D) · 10 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 436 tok/s | Runs great |
| StarCoder 15B | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 tok/s | Runs well | 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 34.9 tok/s | Runs great |
| StarCoder2 3B | 17Basic quality (D) · 17 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 46.1 tok/s | Runs great | 18Basic quality (D) · 18 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 71.8 tok/s | Runs great |
| SmolLM2 360M Instruct | 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 91.0 tok/s | Runs great | 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 279 tok/s | Runs great |
| Aya 23 35B | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 5.6 tok/s | Slowover memory limit | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| StarCoder2 7B | 16Basic quality (D) · 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 32.0 tok/s | Runs great | 23Basic quality (D) · 23 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 41.4 tok/s | Runs great |
| TinyLlama 1.1B Chat | 6Basic quality (D) · 6 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 64.6 tok/s | Runs great | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 91.2 tok/s | Runs great |
| Qwen2 0.5B Instruct | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 66.9 tok/s | Runs great | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 205 tok/s | Runs great |
| ERNIE 4.5 0.3B | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 91.0 tok/s | Runs great | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 279 tok/s | Runs great |
| SmolVLM 256M Instruct | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 126 tok/s | Runs great | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 386 tok/s | Runs great |
| Gemma 3 270M | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 121 tok/s | Runs great | 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 372 tok/s | Runs great |
| DeepSeek Coder 1.3B Instruct | 12Basic quality (D) · 12 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 52.7 tok/s | Runs great | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 74.3 tok/s | Runs great |
| LLM4Decompile 1.3B V2 | 12Basic quality (D) · 12 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 52.7 tok/s | Runs great | 12Basic quality (D) · 12 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 74.3 tok/s | Runs great |
| surya ocr 2.5 | 15Basic quality (D) · 15 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 47.5 tok/s | Runs great | 15Basic quality (D) · 15 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 145 tok/s | Runs great |
| SmolLM 360M Instruct | 5Basic quality (D) · 5 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 91.0 tok/s | Runs great | 5Basic quality (D) · 5 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 279 tok/s | Runs great |
| Falcon 7B Instruct | 4Basic quality (D) · 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 31.8 tok/s | Runs great | 5Basic quality (D) · 5 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 41.1 tok/s | Runs great |
| CodeGemma 2B | 13Basic quality (D) · 13 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 46.0 tok/s | Runs great | 14Basic quality (D) · 14 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. | 86.7 tok/s | Runs great |
| SmolLM2 135M Instruct | 4Basic quality (D) · 4 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 252 tok/s | Runs great | 4Basic quality (D) · 4 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 772 tok/s | Runs great |
| Falcon-H1 Tiny 90M Instruct | 3Basic quality (D) · 3 of 100: how capable the model is at F16, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. | 364 tok/s | Runs great | 3Basic quality (D) · 3 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 1,115 tok/s | Runs great |
| SmolLM 135M Instruct | 2Basic quality (D) · 2 of 100: how capable the model is at F16, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. | 252 tok/s | Runs great | 2Basic quality (D) · 2 of 100: how capable the model is at F16, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. | 772 tok/s | Runs great |
| Qwen2.5 Coder 0.5B Instruct | 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 46/100 (quality, fit, speed and memory use) sets the ranking. | 66.9 tok/s | Runs great | 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. | 205 tok/s | Runs great |
| gpt-oss 120B | 84Great quality (A) · 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 84Great quality (A) · 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K3.1 Pro (2.8T) | 69Good quality (B) · 69 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 69Good quality (B) · 69 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V4.1 Flash 1017 (552B-A16B) | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-5.3 (753B-A40B) | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3.8 Max (2.4T-A95B) | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K3 (2.8T) | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiMo V2.6 Pro (1T-A42B) | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling 3.1 Flash 1019 (560B-A25B) | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 68Good quality (B) · 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax M3.1 Flash (428B-A23B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Step 5 Preview (600B-A27B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-5.3 Flash (320B-A18B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiMo V2.6 Flash (310B-A15B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V4.1 Flash (552B-A16B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mistral Large 4 (1T-A52B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3.8 Flash Next (125B-A6B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-5.2 (753B-A40B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Hunyuan Hy4 Preview (780B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Muse Spark Open (816B-A42B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V4 Pro (1.6T-A49B) | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K2.6 (1T) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling 3.1 Flash (560B-A25B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiMo V2.5 Pro (1T-A42B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Intern-S2 (397B-A17B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-5.1 (744B-A40B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Nex N2 Pro (397B-A17B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V4 Pro (April 2026) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Naive N0.5 Flash (309B-A15.5B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3.5 397B-A17B | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax M3 (427B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-5 (744B-A40B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| K2-Horizon (375B-A23B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Hunyuan Hy3 (295B-A21B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Beam 1.0 (501B-A23B) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K2.5 (1T) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K2.7 Code (1T) | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V4 Flash (284B-A13B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiMo V2.5 (310B-A15B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ornith 1.5 397B-A17B | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Nemotron 3.5 Super VL 120B-A12B | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V3.2 (671B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Inkling Small (276B-A12B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-4.7 (355B-A32B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Inkling (975B-A41B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax M2.7 (230B-A10B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Step 3.7 Flash (198B-A11B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Nemotron 3 Ultra 550B-A55B | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek R1 0528 (671B) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K2 Thinking (1T) | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ornith 1.5 122B-A10B | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V3.2 Exp | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling 2.6 1T | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax M2.5 (230B-A10B) | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Nemotron 3.5 Bolt 74B-A9B | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 19.7 tok/s | Slowheavy RAM offload |
| DeepSeek R1 (671B) | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V4 Flash (April 2026) | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3.5 122B-A10B | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Step 3.5 Flash (196B-A11B) | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V3.1 Terminus | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 64Good quality (B) · 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V3.1 (671B) | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K2 Instruct 0905 (1T) | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kolibri 1 (78B-A3.5B) | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 45.9 tok/s | Slowheavy RAM offload |
| MiniMax M2.1 (230B-A10B) | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3 235B-A22B Thinking 2507 | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3 Coder 480B-A35B | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiMo V2 Flash (309B-A15B) | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling 1T | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mistral Large 3 (675B) | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3 235B-A22B Instruct 2507 | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Solar Open 2 250B | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 63Good quality (B) · 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi K2 Instruct (July 2025) | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax M2 (230B-A10B) | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3-VL 235B-A22B Instruct | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 62Good quality (B) · 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Command A+ (218B-A25B) | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| K-EXAONE 236B-A23B | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| LongCat Flash Chat (560B) | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Nemotron 3 Super 120B-A12B | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3 235B-A22B | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 61Good quality (B) · 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Devstral 2 (123B) | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| InternVL3.5 241B-A28B | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Laguna S 2.1 (118B-A8B) | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Llama 3.1 Nemotron Ultra 253B v1 | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3 Coder Next (80B-A3B) | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3-Next 80B-A3B Thinking | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Baichuan M3 235B-A22B | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-4.6 (355B-A32B) | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen3 Next 80B-A3B Instruct | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V3 0324 (671B) | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-4.6V | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax-M1 80K | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Command A Reasoning (111B) | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-4.5 (355B-A32B) | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling 3.0 Flash (127B-A5B) | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax-M1 40K | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mistral Small 4 (119B-A6B) | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mistral Medium 3.5 (128B) | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Trinity Large Thinking (400B-A13B) | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Command A (111B) | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| ERNIE 4.5 300B-A47B | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-4.5V | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Hermes 4 405B | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kimi Dev 72B | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling 2.6 Flash (104B-A7.4B) | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mistral Large 2.1 (2411) | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Solar Open 100B | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| GLM-4.5 Air (106B-A12B) | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mistral Large 2 (123B) | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| NorthCUA Pro | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Pixtral Large 124B (2411) | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ring Flash 2.0 (100B-A6B) | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| dots.llm1 Instruct | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Ling Flash 2.0 (100B-A6B) | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax-Text-01 | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Tülu 3 405B | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V3 (671B) | 52Fair quality (C) · 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 52Fair quality (C) · 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Sarvam 105B | 52Fair quality (C) · 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 52Fair quality (C) · 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek Coder V2 (236B) | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Hermes 4 70B | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Hunyuan A13B (80B-A13B) | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| InternVL3 78B | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Leanstral 119B-A6B | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Llama 3.3 70B Instruct | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Qwen2.5 72B Instruct | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen2.5-VL 72B Instruct | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 51Fair quality (C) · 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek Coder V2 Instruct (0614) | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek R1 Distill Llama 70B | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| DeepSeek V2.5 1210 | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Llama 3.1 405B Instruct | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Llama 4 Maverick (17B x 128E) | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Llama 4 Scout 17B-16E | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Tülu 3 70B | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Command R+ (08-2024) 104B | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V2.5 | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Groot 65B A6B 261023 | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Jamba 1.5 Large | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Llama 3.1 70B Instruct | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Hermes 3 (Llama 3.1 70B) | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Ling 3.0 flash Sante | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| QVQ 72B Preview | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Apertus v1.5 70B | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. | 2.4 tok/s | Very slowheavy RAM offload |
| Llama 3.1 Nemotron 70B Instruct | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Llama 3.2 90B Vision Instruct | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Command R+ 104B (04-2024) | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen2 72B Instruct | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen2-VL 72B Instruct | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 44Fair quality (C) · 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| DeepSeek V2 Chat | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 43Fair quality (C) · 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Apertus 70B Instruct | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 46/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Llama 3 70B Instruct | 40Fair quality (C) · 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 40Fair quality (C) · 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 46/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Qwen2.5-Math 72B Instruct | 40Fair quality (C) · 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 40Fair quality (C) · 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 41/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 tok/s | Very slowheavy RAM offload |
| Mixtral 8x22B Instruct | 38Basic quality (D) · 38 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 38Basic quality (D) · 38 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen1.5 110B Chat | 36Basic quality (D) · 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 36Basic quality (D) · 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Grok-1 (314B) | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Snowflake Arctic Instruct | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 35Basic quality (D) · 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Code Llama 70B Instruct | 30Basic quality (D) · 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 30Basic quality (D) · 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 36/100 (quality, fit, speed and memory use) sets the ranking. | 2.7 tok/s | Very slowheavy RAM offload |
| DeepSeek LLM 67B Chat | 28Basic quality (D) · 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 28Basic quality (D) · 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 40/100 (quality, fit, speed and memory use) sets the ranking. | 2.8 tok/s | Very slowheavy RAM offload |
| Llama 2 70B Chat | 25Basic quality (D) · 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 25Basic quality (D) · 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 39/100 (quality, fit, speed and memory use) sets the ranking. | 2.7 tok/s | Very slowheavy RAM offload |
| Falcon 180B Chat | 23Basic quality (D) · 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 23Basic quality (D) · 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
Image models
57 of 60 image models run on the AMD EPYC 7002 / 7003 series NAS (32 GB). 26.2 GB usable for models.
54 of 60 image models run on the Intel Arc B580. 12.0 GB usable for models.
| Model | QualityEPYC 7002 / 7003 series NAS | SpeedEPYC 7002 / 7003 series NAS | RunsEPYC 7002 / 7003 series NAS | QualityArc B580 | SpeedArc B580 | RunsArc B580 |
|---|---|---|---|---|---|---|
| Krea 2 Turbo | 90Excellent quality (A+) · 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 2.7 min | Very slow | 90Excellent quality (A+) · 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 21 s | Runs wellCPU offload |
| FLUX.2 [klein] 9B | 82Great quality (A) · 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 67 s | Slow | 78Great quality (A) · 78 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 9.4 s | Runs greatencoder in RAM |
| ERNIE-Image | 87Excellent quality (A+) · 87 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 min | Very slow | 86Excellent quality (A+) · 86 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 86 s | Slowtight fit · encoder in RAM |
| Ideogram 4 | 85Excellent quality (A+) · 85 of 100: how capable the model is at NF4, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 18.8 min | Very slow | 90Excellent quality (A+) · 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 2.0 min | SlowCPU offload |
| FLUX 3 Image Klein | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 4.5 min | Very slow | 84Great quality (A) · 84 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 28 s | Runs well |
| ERNIE-Image Turbo | 83Great quality (A) · 83 of 100: how capable the model is at FP8, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 2.0 min | Slow | 79Great quality (A) · 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 16 s | Runs welltight fit |
| FLUX.1 Krea [dev] | 81Great quality (A) · 81 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 9.5 min | Very slow | 76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 62 s | Slowencoder in RAM |
| Fibo 2.0 Dev | 93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 16.2 min | Very slowencoder unloaded after use | 93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 2.0 min | SlowCPU offload |
| Z-Image Turbo | 82Great quality (A) · 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 1.5 min | Slow | 78Great quality (A) · 78 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 12 s | Runs great |
| FLUX.1 Kontext [dev] | 81Great quality (A) · 81 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 min | Very slow | 76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 1.9 min | Slowtight fit · encoder in RAM |
| FLUX.1 [dev] | 79Great quality (A) · 79 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 9.5 min | Very slow | 74Great quality (A) · 74 of 100: how capable the model is at NF4, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 57 s | Slowencoder in RAM |
| MOSS-Image SFT | 81Great quality (A) · 81 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 11.0 min | Very slow | 80Great quality (A) · 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 55 s | Slowencoder in RAM |
| Qwen-Image 2.1 Turbo | 87Excellent quality (A+) · 87 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 7.5 min | Very slowencoder unloaded after use | 82Great quality (A) · 82 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| FLUX.1 [schnell] | 68Good quality (B) · 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 74 s | Slow | 65Good quality (B) · 65 of 100: how capable the model is at NF4, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 9.5 s | Runs greatencoder in RAM |
| FLUX.2 [klein] 4B | 76Great quality (A) · 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 37 s | Runs well | 76Great quality (A) · 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 3.7 s | Runs great |
| FLUX.1 Fill [dev] | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 7.7 min | Very slow | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 60 s | SlowCPU offload |
| Chroma1-HD | 76Great quality (A) · 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 12.3 min | Very slow | 76Great quality (A) · 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 78 s | Slowtight fit · encoder in RAM |
| FireRed-Image-Edit 1.1 | 80Great quality (A) · 80 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 65.8 min | Very slowtight fit | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 8.2 min | Very slowtight fit · CPU offload |
| Ming-Image 0.1 Design | 84Great quality (A) · 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 10.6 min | Very slowencoder unloaded after use | 84Great quality (A) · 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen-Image-Edit 2511 | 79Great quality (A) · 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 65.8 min | Very slowtight fit | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 8.2 min | Very slowtight fit · CPU offload |
| Z-Image | 81Great quality (A) · 81 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 13.4 min | Very slow | 76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 84 s | Slow |
| Iris-3B | 72Great quality (A) · 72 of 100: how capable the model is at FP16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 11.0 min | Very slow | 72Great quality (A) · 72 of 100: how capable the model is at FP16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 69 s | Slowencoder in RAM |
| HiDream-I1 Full | 83Great quality (A) · 83 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 37.2 min | Very slowencoder unloaded after use | 83Great quality (A) · 83 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 4.6 min | Very slowCPU offload |
| Qwen-Image 2512 | 83Great quality (A) · 83 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 77.7 min | Very slowencoder unloaded after use | 83Great quality (A) · 83 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 9.7 min | Very slowtight fit · CPU offload |
| Stable Diffusion 3.5 Large Turbo | 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 59 s | Slow | 57Good quality (B) · 57 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 8.3 s | Runs greatencoder in RAM |
| CogView4 6B | 68Good quality (B) · 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 11.1 min | Very slow | 68Good quality (B) · 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 70 s | Slowencoder in RAM |
| FLUX.2 [dev] | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 31.1 min | Very slowtight fit · encoder unloaded after use | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 3.9 min | Very slowCPU offload |
| HiDream-I1 Dev | 79Great quality (A) · 79 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 14.1 min | Very slowencoder unloaded after use | 78Great quality (A) · 78 of 100: how capable the model is at FP8, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 82 s | SlowCPU offload |
| OmniGen2 | 66Good quality (B) · 66 of 100: how capable the model is at FP16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 11.0 min | Very slow | 64Good quality (B) · 64 of 100: how capable the model is at FP8, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 59 s | Slowtight fit |
| Stable Diffusion 3.5 Large | 65Good quality (B) · 65 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 8.7 min | Very slow | 64Good quality (B) · 64 of 100: how capable the model is at FP8, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 43 s | Slowtight fit · encoder in RAM |
| HiDream-I1 Fast | 75Great quality (A) · 75 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 8.3 min | Very slowencoder unloaded after use | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 48 s | SlowCPU offload |
| Kolors | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 2.7 min | Very slow | 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 15 s | Runs welltight fit |
| Stable Diffusion 3.5 Medium | 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 min | Very slow | 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 22 s | Runs welltight fit |
| AuraFlow v0.3 | 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 9.6 min | Very slow | 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 52 s | Slowtight fit |
| Stable Diffusion 3 Medium | 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 2.1 min | Slow | 50Fair quality (C) · 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 12 s | Runs great |
| Qwen-Image | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 80.1 min | Very slowCPU offload | 86Excellent quality (A+) · 86 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Lumina-Image 2.0 | 62Good quality (B) · 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 4.8 min | Very slow | 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 26 s | Runs well |
| SDXL-Lightning (4-step) | 54Fair quality (C) · 54 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 29 s | Runs well | 54Fair quality (C) · 54 of 100: how capable the model is at FP16, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 2.9 s | Runs great |
| DreamShaper XL v2 Turbo | 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 53 s | Slow | 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 5.4 s | Runs great |
| PixArt-Σ XL 1024 | 50Fair quality (C) · 50 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 62 s | Slow | 50Fair quality (C) · 50 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 6.5 s | Runs greatencoder in RAM |
| SDXL Turbo | 40Fair quality (C) · 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 6.2 s | Runs great | 40Fair quality (C) · 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 0.6 s | Runs great |
| Juggernaut XL v9 | 62Good quality (B) · 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 3.1 min | Very slow | 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 17 s | Runs well |
| RealVisXL V5.0 | 62Good quality (B) · 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 3.1 min | Very slow | 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 17 s | Runs well |
| Stable Cascade | 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 3.4 min | Very slow | 50Fair quality (C) · 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 19 s | Runs well |
| Animagine XL 4.0 | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 2.9 min | Very slow | 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 16 s | Runs well |
| Illustrious XL v0.1 | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 2.9 min | Very slow | 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 16 s | Runs well |
| Pony Diffusion V6 XL | 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 2.6 min | Very slow | 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 14 s | Runs well |
| Playground v2.5 | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 min | Very slow | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 31 s | Runs well |
| Animagine XL 3.1 | 56Good quality (B) · 56 of 100: how capable the model is at FP16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 2.9 min | Very slow | 54Fair quality (C) · 54 of 100: how capable the model is at FP8, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 16 s | Runs well |
| Stable Diffusion XL 1.0 | 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 3.1 min | Very slow | 53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 17 s | Runs well |
| Hunyuan-DiT v1.2 | 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 4.1 min | Very slow | 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 25 s | Runs well |
| LCM DreamShaper v7 | 30Basic quality (D) · 30 of 100: how capable the model is at FP16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 10 s | Runs great | 30Basic quality (D) · 30 of 100: how capable the model is at FP16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 1.0 s | Runs great |
| SD Turbo | 28Basic quality (D) · 28 of 100: how capable the model is at FP16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 3.3 s | Runs great | 28Basic quality (D) · 28 of 100: how capable the model is at FP16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 0.3 s | Runs great |
| Realistic Vision V5.1 | 42Fair quality (C) · 42 of 100: how capable the model is at FP16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | 49 s | Slow | 42Fair quality (C) · 42 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 s | Runs great |
| DreamShaper 8 | 40Fair quality (C) · 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | 49 s | Slow | 40Fair quality (C) · 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 s | Runs great |
| Kandinsky 3 | 48Fair quality (C) · 48 of 100: how capable the model is at FP16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 6.5 min | Very slowencoder unloaded after use | 48Fair quality (C) · 48 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 39 s | Runs wellencoder in RAM |
| Stable Diffusion 1.5 | 35Basic quality (D) · 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 49 s | Slow | 35Basic quality (D) · 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 s | Runs great |
| Cosmos3-Super Text2Image (64B) | 90Excellent quality (A+) · 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 90Excellent quality (A+) · 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| HunyuanImage 3.0 (80B-A13B) | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Qwen-Image 2.1 | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
Video models
37 of 43 video models run on the AMD EPYC 7002 / 7003 series NAS (32 GB). 26.2 GB usable for models.
14 of 43 video models run on the Intel Arc B580. 12.0 GB usable for models.
| Model | QualityEPYC 7002 / 7003 series NAS | SpeedEPYC 7002 / 7003 series NAS | RunsEPYC 7002 / 7003 series NAS | QualityArc B580 | SpeedArc B580 | RunsArc B580 |
|---|---|---|---|---|---|---|
| Bernini-R A14B (video generation + editing) | 83Great quality (A) · 83 of 100: how capable the model is at FP8, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 7.8 h | Very slowencoder unloaded after use | 83Great quality (A) · 83 of 100: how capable the model is at FP8, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 58.0 min | Very slowtight fit · CPU offload |
| HY-OmniWeaving (8.3B, unified video) | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 5.4 h | Very slowtight fit | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 min | Very slowCPU offload |
| Wan 2.2 S2V 14B (speech to video) | 70Great quality (A) · 70 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 13.7 h | Very slow | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 80.5 min | Very slowtight fit · CPU offload |
| Wan 2.1 I2V 14B 480P | 68Good quality (B) · 68 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 12.3 h | Very slow | 72Great quality (A) · 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 72.5 min | Very slowtight fit · CPU offload |
| LTX-2.3 | 79Great quality (A) · 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 5.8 h | Very slowtight fit · encoder unloaded after use | 79Great quality (A) · 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| FramePack I2V (HunyuanVideo 13B) | 72Great quality (A) · 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 4.3 h | Very slowencoder unloaded after use | 72Great quality (A) · 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 32.2 min | Slowtight fit · CPU offload |
| SCAIL-2 14B (character animation) | 78Great quality (A) · 78 of 100: how capable the model is at FP8, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 12.9 h | Very slowtight fit · encoder unloaded after use | 78Great quality (A) · 78 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| LTX-2.5 (22B, video + audio) | 87Excellent quality (A+) · 87 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 1.6 h | Very slowCPU offload | 81Great quality (A) · 81 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| LTX-Video 2B | 50Fair quality (C) · 50 of 100: how capable the model is at FP16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 17.1 min | Slow | 49Fair quality (C) · 49 of 100: how capable the model is at FP8, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | 1.5 min | Runs welltight fit |
| LTX-2 (19B, video + audio) | 76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 5.8 h | Very slowtight fit · encoder unloaded after use | 76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| LTX-Video 13B 0.9.8 (distilled) | 66Good quality (B) · 66 of 100: how capable the model is at FP8, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 38.8 min | Slowtight fit · encoder unloaded after use | 66Good quality (B) · 66 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.1 T2V 1.3B | 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 2.7 h | Very slow | 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 17.3 min | Slowencoder in RAM |
| Wan 2.1 VACE 1.3B | 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 3.0 h | Very slow | 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 18.8 min | Slowencoder in RAM |
| SkyReels V2 T2V 14B 540P | 72Great quality (A) · 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 15.1 h | Very slowtight fit · encoder unloaded after use | 72Great quality (A) · 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| MiniMax H3 (33B, video + audio) | 89Excellent quality (A+) · 89 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 17.7 h | Very slowtight fit · CPU offload | 83Great quality (A) · 83 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kandinsky 6.0 Video Lite (3B, video + audio) | 68Good quality (B) · 68 of 100: how capable the model is at FP16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 3.7 h | Very slowencoder unloaded after use | 68Good quality (B) · 68 of 100: how capable the model is at FP16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 27.7 min | SlowCPU offload |
| CogVideoX 5B I2V | 53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 1.8 h | Very slow | 53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 11.7 min | Runs wellencoder in RAM |
| SkyReels V2 Diffusion Forcing 1.3B | 50Fair quality (C) · 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 3.1 h | Very slow | 50Fair quality (C) · 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | 19.5 min | Slowtight fit · encoder in RAM |
| CogVideoX 5B | 53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 1.8 h | Very slow | 53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 11.7 min | Runs wellencoder in RAM |
| Wan 2.2 I2V A14B | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 27.3 h | Very slowtight fit · CPU offload | 79Great quality (A) · 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.2 T2V A14B | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 27.3 h | Very slowtight fit · CPU offload | 79Great quality (A) · 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.2 TI2V 5B | 65Good quality (B) · 65 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 7.5 h | Very slowtight fit · encoder unloaded after use | 65Good quality (B) · 65 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Mochi 1 Preview (10B) | 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 11.9 h | Very slowencoder unloaded after use | 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| HunyuanVideo 1.5 (8.3B) | 82Great quality (A) · 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 24.0 h | Very slowtight fit · CPU offload | 82Great quality (A) · 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Cosmos-Predict2 2B Video2World | 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | 2.7 h | Very slowencoder unloaded after use | 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.2 Animate 14B | 77Great quality (A) · 77 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 32.4 h | Very slowtight fit · CPU offload | 72Great quality (A) · 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.1 I2V 14B 720P | 77Great quality (A) · 77 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 34.1 h | Very slowtight fit · CPU offload | 72Great quality (A) · 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.1 T2V 14B | 77Great quality (A) · 77 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 34.1 h | Very slowtight fit · CPU offload | 72Great quality (A) · 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan 2.1 VACE 14B | 77Great quality (A) · 77 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | 37.5 h | Very slowtight fit · CPU offload | 72Great quality (A) · 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Step-Video T2V (30B) | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 41.4 h | Very slowtight fit · CPU offload | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Open-Sora 2.0 (11B) | 70Great quality (A) · 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | 44.5 h | Very slowCPU offload | 70Great quality (A) · 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| CogVideoX 2B | 44Fair quality (C) · 44 of 100: how capable the model is at FP8, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | 37.0 min | Slow | 44Fair quality (C) · 44 of 100: how capable the model is at FP8, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 3.7 min | Runs welltight fit |
| Cosmos-Predict2 14B Video2World | 68Good quality (B) · 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | 21.4 h | Very slowtight fit · CPU offload | 68Good quality (B) · 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| HunyuanVideo 13B | 66Good quality (B) · 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 h | Very slowtight fit · CPU offload | 66Good quality (B) · 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| HunyuanVideo I2V 13B | 66Good quality (B) · 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | 42.7 h | Very slowtight fit · CPU offload | 66Good quality (B) · 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| AnimateDiff v3 (SD 1.5) | 35Basic quality (D) · 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 13.7 min | Slow | 35Basic quality (D) · 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 83 s | Runs great |
| Allegro (2.8B) | 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. | 17.8 h | Very slowtight fit · CPU offload | 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Cosmos3-Super Image2Video (64B) | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Kandinsky 6.0 Pro (29B, video + audio) | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| SkyReels V2 T2V 14B 720P | 76Great quality (A) · 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 76Great quality (A) · 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Wan-Dancer 14B (music to dance) | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 74Great quality (A) · 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Cosmos 1.0 Diffusion 7B Text2World | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| CogVideoX 1.5 5B | 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
Decision models
67 of 70 decision models run on the AMD EPYC 7002 / 7003 series NAS (32 GB). 26.2 GB usable for models.
68 of 70 decision models run on the Intel Arc B580. 12.0 GB usable for models.
| Model | QualityEPYC 7002 / 7003 series NAS | SpeedEPYC 7002 / 7003 series NAS | RunsEPYC 7002 / 7003 series NAS | QualityArc B580 | SpeedArc B580 | RunsArc B580 |
|---|---|---|---|---|---|---|
| OpenAI Privacy Filter | 78Great quality (A) · 78 of 100: how capable the model is at F16, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 44 ms | Runs great | 78Great quality (A) · 78 of 100: how capable the model is at F16, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 5.1 ms | Runs great |
| Llama Prompt Guard 2 22M | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 103 ms | Runs great | 68Good quality (B) · 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 7.0 ms | Runs great |
| GLiNER2.5 Decide | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 345 ms | Runs well | 80Great quality (A) · 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 14 ms | Runs great |
| Jina Reranker v3.5 | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 780 ms | Slow | 86Excellent quality (A+) · 86 of 100: how capable the model is at F16, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. | 30 ms | Runs great |
| Jina Reranker v3 | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 780 ms | Slow | 84Great quality (A) · 84 of 100: how capable the model is at F16, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 30 ms | Runs great |
| Qwen3 Reranker 4B | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 5.2 s | Very slow | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 160 ms | Runs well |
| GLiClass Base v3 | 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 136 ms | Runs well | 66Good quality (B) · 66 of 100: how capable the model is at F16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 8.0 ms | Runs great |
| FRIDA-Decisions | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 574 ms | Slow | 80Great quality (A) · 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 21 ms | Runs great |
| GLiClass ModernBERT Large v3 | 74Great quality (A) · 74 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 282 ms | Runs well | 74Great quality (A) · 74 of 100: how capable the model is at F16, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| DeBERTa v3 Base Zero-shot v2.0 | 64Good quality (B) · 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 129 ms | Runs great | 64Good quality (B) · 64 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 7.8 ms | Runs great |
| Qwen3 Reranker 8B | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 10.6 s | Very slow | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | 321 ms | Runs well |
| GTE Reranker ModernBERT Base | 70Great quality (A) · 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 213 ms | Runs well | 70Great quality (A) · 70 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 10 ms | Runs great |
| Llama Prompt Guard 2 86M | 76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 394 ms | Runs well | 76Great quality (A) · 76 of 100: how capable the model is at F16, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 16 ms | Runs great |
| zerank-2 | 88Excellent quality (A+) · 88 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 5.2 s | Very slow | 88Excellent quality (A+) · 88 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 160 ms | Runs well |
| Skywork Reward V2 Llama 3.1 8B | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 20.9 s | Very slow | 90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 560 ms | Slowtight fit |
| Qwen3 Reranker 0.6B | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 780 ms | Slow | 80Great quality (A) · 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 30 ms | Runs great |
| DeBERTa v3 Large Zero-shot v2.0 | 72Great quality (A) · 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 310 ms | Runs well | 72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| Granite Embedding Reranker English R2 | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 213 ms | Runs well | 68Good quality (B) · 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 10 ms | Runs great |
| GLiNER2 Base v1 | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 150 ms | Runs well | 68Good quality (B) · 68 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 8.5 ms | Runs great |
| Shieldstral 1.0 3B | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 s | Very slow | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 153 ms | Runs well |
| Intern-Decision 4B | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 3.0 s | Very slow | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. | 102 ms | Runs great |
| ModernBERT Large Zero-shot v2.0 | 70Great quality (A) · 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 282 ms | Runs well | 70Great quality (A) · 70 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| mxbai-rerank-large v2 | 82Great quality (A) · 82 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 2.0 s | Very slow | 82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 72 ms | Runs great |
| d1 3B | 82Great quality (A) · 82 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 2.0 s | Very slow | 82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 88 ms | Runs great |
| mxbai-rerank-base v2 | 76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 638 ms | Slow | 76Great quality (A) · 76 of 100: how capable the model is at F16, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | 26 ms | Runs great |
| GLiGuard 300M | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 296 ms | Runs well | 68Good quality (B) · 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| Laya | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 296 ms | Runs well | 68Good quality (B) · 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| zerank-1 small | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 2.2 s | Very slow | 80Great quality (A) · 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 80 ms | Runs great |
| ProtectAI DeBERTa v3 Prompt Injection v2 | 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 254 ms | Runs well | 66Good quality (B) · 66 of 100: how capable the model is at F16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 12 ms | Runs great |
| Jina Reranker v2 Base Multilingual | 70Great quality (A) · 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 394 ms | Runs well | 70Great quality (A) · 70 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. | 16 ms | Runs great |
| Skywork Reward V2 Qwen3 8B | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 21.1 s | Very slow | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | 565 ms | Slowtight fit |
| LemonDecider 4B | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 3.2 s | Very slow | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. | 88 ms | Runs great |
| ms-marco MiniLM L6 v2 | 52Fair quality (C) · 52 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 41 ms | Runs great | 52Fair quality (C) · 52 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 5.0 ms | Runs great |
| DuoGuard 0.5B | 72Great quality (A) · 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 686 ms | Slow | 72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 25 ms | Runs great |
| BGE Reranker v2 Gemma | 78Great quality (A) · 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 3.3 s | Very slow | 78Great quality (A) · 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | 101 ms | Runs great |
| BGE Reranker v2 M3 | 72Great quality (A) · 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 797 ms | Slow | 72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 28 ms | Runs great |
| Qwen3Guard Gen 8B | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 11.2 s | Very slow | 81Great quality (A) · 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 398 ms | Runs well |
| Laya Multilingual | 60Good quality (B) · 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 227 ms | Runs well | 60Good quality (B) · 60 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 11 ms | Runs great |
| Qwen3Guard Gen 4B | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 6.0 s | Very slow | 83Great quality (A) · 83 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | 244 ms | Runs well |
| HHEM 2.1 Open | 62Good quality (B) · 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 310 ms | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| GLiNER Multi v2.1 | 62Good quality (B) · 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 206 ms | Runs well | 62Good quality (B) · 62 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 10 ms | Runs great |
| Llama 3.1 Nemotron Safety Guard 8B v3 | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 11.0 s | Very slow | 79Great quality (A) · 79 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 390 ms | Runs well |
| Granite Guardian 4.1 8B | 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 11.4 s | Very slow | 83Great quality (A) · 83 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | 459 ms | Slow |
| Skywork Reward V2 Qwen3 0.6B | 72Great quality (A) · 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 1.7 s | Very slow | 72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | 55 ms | Runs great |
| Qwen3Guard Gen 0.6B | 74Great quality (A) · 74 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 1.0 s | Slow | 74Great quality (A) · 74 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | 74 ms | Runs great |
| NuExtract 2.0 4B | 76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 3.7 s | Very slow | 75Great quality (A) · 75 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 371 ms | Runs well |
| WildGuard 7B | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 9.9 s | Very slow | 79Great quality (A) · 79 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 398 ms | Runs well |
| Granite Guardian 3.3 8B | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 11.2 s | Very slow | 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | 397 ms | Runs well |
| BART Large MNLI | 56Good quality (B) · 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | 289 ms | Runs well | 56Good quality (B) · 56 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 13 ms | Runs great |
| NuExtract 2.0 2B | 70Great quality (A) · 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 2.2 s | Very slow | 66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | 166 ms | Runs well |
| LettuceDetect Large | 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 1.1 s | Slow | 66Good quality (B) · 66 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | 38 ms | Runs great |
| DynaGuard 8B | 79Great quality (A) · 79 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 28.9 s | Very slow | 79Great quality (A) · 79 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 6.3 s | Very slow |
| Nemotron Content Safety Reasoning 4B | 78Great quality (A) · 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 15.2 s | Very slow | 77Great quality (A) · 77 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. | 2.5 s | Very slow |
| gpt-oss-safeguard 20B | 76Great quality (A) · 76 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 9.0 s | Very slow | 76Great quality (A) · 76 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | 2.6 s | Very slowoffloads to RAM |
| Selene 1 Mini 8B | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 37.5 s | Very slow | 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 6.1 s | Very slow |
| CompassJudger-2 32B | 81Great quality (A) · 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 119.1 s | Very slowtight fit | 85Excellent quality (A+) · 85 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 63.7 s | Very slowheavy RAM offload |
| ShieldGemma 9B | 64Good quality (B) · 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 12.0 s | Very slow | 63Good quality (B) · 63 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | 350 ms | Runs well |
| CompassJudger-2 7B | 78Great quality (A) · 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 35.6 s | Very slow | 78Great quality (A) · 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 5.8 s | Very slow |
| ShieldGemma 27B | 65Good quality (B) · 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 35.1 s | Very slowtight fit | 65Good quality (B) · 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. | 1.8 s | Very slowheavy RAM offload |
| Llama Guard 3 8B | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | 11.0 s | Very slow | 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | 390 ms | Runs well |
| ShieldGemma 2B | 58Good quality (B) · 58 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 3.4 s | Very slow | 58Good quality (B) · 58 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | 105 ms | Runs great |
| DynaGuard 1.7B | 70Great quality (A) · 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | 7.2 s | Very slow | 69Good quality (B) · 69 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | 1.2 s | Slow |
| Llama Guard 4 12B | 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | 16.4 s | Very slow | 64Good quality (B) · 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | 617 ms | Slow |
| Llama Guard 3 1B | 56Good quality (B) · 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 2.1 s | Very slow | 56Good quality (B) · 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | 96 ms | Runs great |
| Prometheus 2 7B | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 33.8 s | Very slow | 68Good quality (B) · 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 5.5 s | Very slow |
| GLIDER | 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | 17.8 s | Very slow | 65Good quality (B) · 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | 2.2 s | Very slow |
| Flow Judge v0.1 | 64Good quality (B) · 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | 17.8 s | Very slow | 63Good quality (B) · 63 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | 2.2 s | Very slow |
| GLiDE 27B | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| gpt-oss-safeguard 120B | 80Great quality (A) · 80 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 80Great quality (A) · 80 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run |
| Prometheus 2 8x7B | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. | — | Won't run | 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. | 18.7 s | Very slowheavy RAM offload |
Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.