Image models · Alibaba
Qwen-Image 2512
- Backbone
- 20.4B
- Native output
- 1328×1328
- Default steps
- 50
- License
- apache-2.0
- Released
- 2025-12
How to run it
Quantized files · GGUF for ComfyUI
Smaller GGUF versions for ComfyUI (with the ComfyUI-GGUF node). No sign-in needed; the model’s license still applies.
huggingface.co/unsloth/Qwen-Image-2512-GGUF (opens in a new tab)
Original weights · Hugging Face
Full-precision files from Alibaba, for fine-tuning, vLLM, Transformers, Diffusers or ComfyUI.
Official repo · GitHub
Reference code, inference scripts and docs from Alibaba. The weights themselves download from Hugging Face.
github.com/QwenLM/Qwen-Image (opens in a new tab)
git clone https://github.com/QwenLM/Qwen-ImageMemory needed
Peak GPU memory at the native output size. Offloading moves the text encoder, or the whole pipeline, to system RAM between steps.
| Precision | Everything on GPU | Text encoder offloaded | Sequential offload | Download |
|---|---|---|---|---|
FP16 | 63.8 GB | 47.2 GB | 20.4 GB | Original weights (opens in a new tab) |
FP8 | 35.1 GB | 26.8 GB | 13.3 GB | Original weights (opens in a new tab) |
Q8 GGUF | 36.5 GB | 28.2 GB | 13.8 GB | GGUF (opens in a new tab) |
Q4 GGUF | 26.9 GB | 18.6 GB | 10.4 GB | GGUF (opens in a new tab) |
Components
- Backbone: 20.43B
- VAE: 0.127B
- Qwen2.5-VL 7B: 8.29B
Best hardware for Qwen-Image 2512
The least expensive GPU in our catalog that runs it without offloading is the Intel Arc B580 (launch price $249) at Q4 GGUF.
| Grade · points | Hardware | Precision | Memory | Time per output | Fit |
|---|---|---|---|---|---|
| 86 ptsExcellent (S) · 86 of 100 points: Model quality 40.5/45, Memory fit 25.0/25, Speed 5.8/15, Memory use 15.0/15 | NVIDIA RTX PRO 6000 Blackwell | FP16 | 63.8 GB / 96.0 GB | 66 s | Runs well |
| 84 ptsGreat (A) · 84 of 100 points: Model quality 40.5/45, Memory fit 25.0/25, Speed 3.6/15, Memory use 15.0/15 | AMD Ryzen AI Max+ 395 (128 GB) | FP16 | 63.8 GB / 96.0 GB | 2.9 min | Runs well |
| 82 ptsGreat (A) · 82 of 100 points: Model quality 37.7/45, Memory fit 25.0/25, Speed 5.0/15, Memory use 15.0/15 | NVIDIA GeForce RTX 5090 | Q4 GGUF | 26.9 GB / 32.0 GB | 1.5 min | Runs well |
| 82 ptsGreat (A) · 82 of 100 points: Model quality 40.5/45, Memory fit 25.0/25, Speed 2.1/15, Memory use 15.0/15 | Apple M4 Max (128 GB) | FP16 | 63.8 GB / 96.0 GB | 8.8 min | Runs well |
| 82 ptsGreat (A) · 82 of 100 points: Model quality 40.1/45, Memory fit 25.0/25, Speed 1.9/15, Memory use 15.0/15 | Apple M4 Max (64 GB) | Q8 GGUF | 36.5 GB / 48.0 GB | 10.1 min | Runs well |
| 81 ptsGreat (A) · 81 of 100 points: Model quality 40.5/45, Memory fit 25.0/25, Speed 3.5/15, Memory use 12.8/15 | NVIDIA GB10 (128 GB) | FP16 | 63.8 GB / 115 GB | 3.1 min | Runs well |
| 81 ptsGreat (A) · 81 of 100 points: Model quality 40.1/45, Memory fit 25.0/25, Speed 1.6/15, Memory use 15.0/15 | Apple M1 Max (64 GB) | Q8 GGUF | 36.5 GB / 48.0 GB | 15.5 min | Runs well |
| 79 ptsGreat (A) · 79 of 100 points: Model quality 37.7/45, Memory fit 25.0/25, Speed 1.5/15, Memory use 15.0/15 | Apple M4 Pro (48 GB) | Q4 GGUF | 26.9 GB / 36.0 GB | 17.7 min | Runs well |
| 78 ptsGreat (A) · 78 of 100 points: Model quality 40.5/45, Memory fit 25.0/25, Speed 2.4/15, Memory use 10.2/15 | Apple M2 Ultra (192 GB) | FP16 | 63.8 GB / 144 GB | 6.6 min | Runs well |
| 77 ptsGreat (A) · 77 of 100 points: Model quality 37.7/45, Memory fit 20.0/25, Speed 4.5/15, Memory use 15.0/15 | NVIDIA GeForce RTX 4090 | Q4 GGUF | 18.6 GB / 24.0 GB | 1.9 min | Runs well |
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.