Image models · Alibaba

Qwen-Image 2.1

Name: Qwen-Image 2.1
Author: Alibaba

Backbone: 7.1B
Native output: 2048×2048
Default steps: 40
License: qwen-research
Released: 2026-09

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-2.1-GGUF (opens in a new tab)

Original weights · Hugging Face

Full-precision files from Alibaba, for fine-tuning, vLLM, Transformers, Diffusers or ComfyUI.

huggingface.co/Qwen/Qwen-Image-2.1 (opens in a new tab)

No sign-in needed.

Official repo · GitHub

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

github.com/QwenLM/Qwen-Image-2.1 (opens in a new tab)

git clone https://github.com/QwenLM/Qwen-Image-2.1

Links checked 1 day ago against Hugging Face, Ollama and GitHub.

Apps that run Qwen-Image 2.1

Our pick for each kind of computer, assuming the model fits in its memory. Pick your own GPU or device in the Model Finder for an exact answer.

Mac (Apple silicon)

Draw Things

Native Mac app with its own fast Metal engine and quantized downloads.

Also good:

Windows PC, NVIDIA GPU

ComfyUI

Runs this model with built-in or popular custom nodes; GGUF files load with ComfyUI-GGUF.

Also good:

Linux, NVIDIA or AMD GPU

ComfyUI

Runs this model with built-in or popular custom nodes; GGUF files load with ComfyUI-GGUF.

Also good:

App	Type	Opens it as	Systems	Engine	Start
Unsloth Desktop	Desktop app	GGUF	macOSWindowsLinux	llama.cpp, MLX, stable-diffusion.cpp	Open in Unsloth Desktop
ComfyUI	Desktop app	Original or GGUF	macOSWindowsLinux	PyTorch, GGUF via ComfyUI-GGUF	Get ComfyUI
Draw Things	Desktop app	Built-in support	macOSiOS	Draw Things	Open in Draw Things
SwarmUI	Desktop app	Original or GGUF	macOSWindowsLinux	ComfyUI	Get SwarmUI
Forge Neo	Desktop app	Built-in support	WindowsLinux	PyTorch	Get Forge Neo
SD.Next	Desktop app	Original or GGUF	macOSWindowsLinux	PyTorch	Get SD.Next
Stability Matrix	Desktop app	Original or GGUF	macOSWindowsLinux	installs ComfyUI, Forge, InvokeAI, SwarmUI, SD.Next or Fooocus	Get Stability Matrix

Which apps can open a model depends on the files published for it: GGUF for llama.cpp-based apps, MLX for Apple’s engine, the original weights for vLLM and SGLang. Compare every app.

Memory 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`	57.0 GB	39.5 GB	29.5 GB	Original weights (opens in a new tab)
`FP8`	41.1 GB	32.3 GB	27.0 GB	Original weights (opens in a new tab)
`Q8 GGUF`	41.6 GB	32.8 GB	27.2 GB	GGUF (opens in a new tab)
`Q4 GGUF`	38.3 GB	29.5 GB	26.0 GB	GGUF (opens in a new tab)

Components

Backbone: 7.12B
VAE: 0.338B
Qwen3-VL 8B: 8.77B

Best hardware for Qwen-Image 2.1

The least expensive GPU in our catalog that runs it without offloading is the Intel Arc Pro B70 (launch price $949) at Q4 GGUF.

Quality	Hardware	Quant	Memory	Speed	Runs
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA RTX PRO 6000 Blackwell	`FP8`	41.1 GB / 96.0 GB	40 s	Runs well
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA H100 80GB SXM	`FP8`	41.1 GB / 80.0 GB	38 s	Runs well
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA H800 80GB SXM	`FP8`	41.1 GB / 80.0 GB	38 s	Runs well
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA GH200 Grace Hopper (96 GB GPU)	`FP8`	41.1 GB / 96.0 GB	38 s	Runs well
93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA B200	`FP16`	57.0 GB / 180 GB	37 s	Runs well
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA H200 141GB SXM	`FP8`	41.1 GB / 141 GB	38 s	Runs well
93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking.	AMD Ryzen AI Max+ 395 (128 GB)	`FP16`	57.0 GB / 96.0 GB	1.9 min	Slow
93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA RTX Spark N1X (6144-core GPU) (128 GB)	`FP16`	57.0 GB / 102 GB	2.2 min	Slow
93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA GB10 (128 GB)	`FP16`	57.0 GB / 115 GB	2.1 min	Slow
86Excellent quality (A+) · 86 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking.	NVIDIA GeForce RTX 5090	`Q4 GGUF`	29.5 GB / 32.0 GB	63 s	Slowtight fit · encoder in RAM

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.