Check my GPU

LLMs ·

Llama 3.3 Nemotron Super 49B v1.5

Parameters
50B
Architecture
Dense
Max context
128K tokens
License
nvidia-open-model-license
Released
2025-07

How to run it

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

huggingface.co/unsloth/Llama-3_3-Nemotron-Super-49B-v1_5-GGUF (opens in a new tab)

ollama run hf.co/unsloth/Llama-3_3-Nemotron-Super-49B-v1_5-GGUF

Original weights · Hugging Face

Full-precision files from NVIDIA, for fine-tuning, vLLM, Transformers.

huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from NVIDIA. The weights themselves download from Hugging Face.

github.com/NVIDIA-NeMo/Nemotron (opens in a new tab)

git clone https://github.com/NVIDIA-NeMo/Nemotron

NVIDIA’s Nemotron repo with recipes for the whole family.

Links checked 6 days ago against Hugging Face, Ollama and GitHub.

Memory needed

Weights plus KV cache and runtime overhead. Leave a little extra free for the operating system and other apps.

KV cache: pruned Llama: some layers skip attention, so the estimate is on the safe side.

Quantization Quality Weights Total at 8K context Total at 32K context Download
Q2_K Low 19.9 GB 23.2 GB 31.3 GB GGUF (opens in a new tab)
Q3_K_M Low 24.4 GB 27.7 GB 35.8 GB GGUF (opens in a new tab)
Q4_K_M Good 30.4 GB 33.7 GB 41.8 GB GGUF (opens in a new tab)
Q5_K_M High 35.4 GB 38.7 GB 46.7 GB GGUF (opens in a new tab)
Q6_K High 40.9 GB 44.2 GB 52.2 GB GGUF (opens in a new tab)
Q8_0 Very high 52.9 GB 56.1 GB 64.2 GB GGUF · folder (opens in a new tab)
F16 Full 99.7 GB 103 GB 111 GB GGUF · folder (opens in a new tab) Original weights (opens in a new tab)

Best hardware for Llama 3.3 Nemotron Super 49B v1.5

The least expensive GPU in our catalog that runs it without offloading is the Intel Arc Pro B70 (launch price $949) at Q3_K_M.

Grade · points Hardware Quant Memory Speed Fit
84 ptsGreat (A) · 84 of 100 points: Model quality 34.9/45, Memory fit 25.0/25, Speed 14.3/15, Memory use 10.5/15 NVIDIA RTX PRO 6000 Blackwell Q5_K_M 43.5 GB / 96.0 GB 36.4 tok/s Runs well
84 ptsGreat (A) · 84 of 100 points: Model quality 30.2/45, Memory fit 24.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 5090 Q3_K_M 29.3 GB / 32.0 GB 57.0 tok/s Tight fit
82 ptsGreat (A) · 82 of 100 points: Model quality 34.2/45, Memory fit 25.0/25, Speed 8.1/15, Memory use 15.0/15 Apple M4 Max (64 GB) Q4_K_M 37.5 GB / 48.0 GB 11.6 tok/s Runs well
81 ptsGreat (A) · 81 of 100 points: Model quality 34.2/45, Memory fit 25.0/25, Speed 6.9/15, Memory use 15.0/15 Apple M1 Max (64 GB) Q4_K_M 37.5 GB / 48.0 GB 8.5 tok/s Runs well
79 ptsGreat (A) · 79 of 100 points: Model quality 35.6/45, Memory fit 25.0/25, Speed 6.7/15, Memory use 12.5/15 Apple M4 Max (128 GB) Q6_K 51.9 GB / 96.0 GB 8.0 tok/s Runs well
77 ptsGreat (A) · 77 of 100 points: Model quality 35.6/45, Memory fit 25.0/25, Speed 4.6/15, Memory use 12.5/15 AMD Ryzen AI Max+ 395 (128 GB) Q6_K 51.9 GB / 96.0 GB 3.8 tok/s Runs well
76 ptsGreat (A) · 76 of 100 points: Model quality 35.6/45, Memory fit 25.0/25, Speed 5.1/15, Memory use 10.7/15 NVIDIA GB10 (128 GB) Q6_K 53.4 GB / 115 GB 4.7 tok/s Runs well
76 ptsGreat (A) · 76 of 100 points: Model quality 34.2/45, Memory fit 25.0/25, Speed 9.8/15, Memory use 7.2/15 Apple M2 Ultra (192 GB) Q4_K_M 45.2 GB / 144 GB 17.0 tok/s Runs well
75 ptsGreat (A) · 75 of 100 points: Model quality 30.2/45, Memory fit 24.0/25, Speed 6.5/15, Memory use 15.0/15 Apple M4 Pro (48 GB) Q3_K_M 30.6 GB / 36.0 GB 7.5 tok/s Tight fit
74 ptsGreat (A) · 74 of 100 points: Model quality 34.2/45, Memory fit 25.0/25, Speed 9.9/15, Memory use 5.9/15 Apple M3 Ultra (256 GB) Q4_K_M 49.1 GB / 192 GB 17.4 tok/s Runs well

Check it on your own hardware

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.