• It's a heck of a lot smaller than Qwen-Image 1 (20b parameters) at only 7b, making it one of the smaller open-weight models available (Z-Image Turbo is one of the few that is smaller at 6b) when compared to Ideogram, Krea2, Flux2, etc.
• It supports native transparency (Qwen's team, as far as I know, is the only one attempting to tackle this). Even though it's relatively trivial to set up background removal postprocessors, it's also neat to see it natively supported.
• It's fast using QwenImage2.1 convrot, a 1MP image took around ~5 seconds on an RTX4090.
Negatives
• The license (assuming you respect it) is far more restrictive. The original Qwen Image 1 was released under the standard Apache license; this one explicitly forbids commercial usage without obtaining a separate license. On the other hand, a lot of us didn't expect the Qwen team to ever release "weights-available" ever again.
Qwen-Image 1.0, released about a year ago, only scored 4/15 on my GenAI Showdown Benchmarks. Since that time, they've been upstaged by Krea 2 (6/15) and Ideogram4 (8/15). I'll post the new results once I have some more time to run them.
I run a prompt-to-ui design site that uses image models for the design process[1]. The text rendering especially makes this model deeply interesting to me, despite the license. Here are some tests using my harness comparing the outputs of gpt-image-2 and qwen 2.1:
The text rendering definitely is much, much better than anything else on the open weights market right now. Small text fidelity is quite good. It seems like the text encoder however gets a little bit overloaded with larger prompts - note the presence of hex codes in the design output, those were inputs from the expanded prompt.
I'll be trying a post-training run on this for web design, it has some serious potential.
The capabilities of local LLM text-to-image is honestly pretty damn impressive. IMO, I think local image generation is currently ahead of local code generation. I can get an image in seconds locally with the quality being way higher than what I'd expect from a local model. However with coding it's much slower and much less impressive. I'm sure there's a reason for this and I'm not an AI expert so I'll let the smarter folks tell me why, but that's just been my observation thus far.
The minute you start looking at other non-frontier models, you understand why Dario, Altman, and Musk are coming together to say we need regulation.
None are as good yet, but what everyone said is coming true - models are not moats. And these folks need an exit (even Msuk whose shares are still locked)
I am really grateful to the Chinese Labs for open sourcing their best models. If it was left to the Americans, we would be forced to pay obscene API fees to use them.
Anyone get success editing videos, frame by frame, using image editing AI? Which model works well?
In my experience, video models generate videos pretty well but are mid at editing. They actually regenerate the entire video along with the edit. So these models being non-deterministic tweak the rest of the video as well, the parts you hoped would be left not edited. It gets exponentially worse when there are humans in the videos, annoying face distortions and for some reason these models just don't understand fingers.
Its happy to see a new open image model from qwen. But the license is a let down. And it dosent even beat their closed qwen3 image wich is already a bit old.
My first impression is that it's not so good at following prompt directions. I asked it to place a 3D text made of glass in a particular city. It instead gave me a broken 3D text on a white background. Maybe with different seeds it gets better, but it's more of a trial and error process than reliable results.
Qwen and Alibaba are the biggest competitor for basically every model out there. They're beating the benchmarks like top-frontier models, focused on open-source and much cheaper than the competitors.
I'm actually looking for a model that is able to create old school pixel art (like 90s style, games like Sierra and so on). But to this date, I haven't seen anything yet. It all reeks of "AI slop". Maybe this is a good thing, I don't know. But if anyone has any tips, I'd appreciate. It's for my own personal use.
Not sure there's a better avatar for the absurdity of AI slop imagery than the "cowboy on horseback". That's a pony with a child's saddle on it, and they've composited a grown man on top of it.
Qwen Image 2.1
(qwen.ai)728 points by jmillikin 20 September 2026 | 196 comments
Comments
Positives
• It's a heck of a lot smaller than Qwen-Image 1 (20b parameters) at only 7b, making it one of the smaller open-weight models available (Z-Image Turbo is one of the few that is smaller at 6b) when compared to Ideogram, Krea2, Flux2, etc.
• It supports native transparency (Qwen's team, as far as I know, is the only one attempting to tackle this). Even though it's relatively trivial to set up background removal postprocessors, it's also neat to see it natively supported.
• It's fast using QwenImage2.1 convrot, a 1MP image took around ~5 seconds on an RTX4090.
Negatives
• The license (assuming you respect it) is far more restrictive. The original Qwen Image 1 was released under the standard Apache license; this one explicitly forbids commercial usage without obtaining a separate license. On the other hand, a lot of us didn't expect the Qwen team to ever release "weights-available" ever again.
Qwen-Image 1.0, released about a year ago, only scored 4/15 on my GenAI Showdown Benchmarks. Since that time, they've been upstaged by Krea 2 (6/15) and Ideogram4 (8/15). I'll post the new results once I have some more time to run them.
https://genai-showdown.specr.net
https://en.wikipedia.org/wiki/Qwen#List_of_models
Unfortunately, it looks like this model is using a much more restrictive license:
https://github.com/QwenLM/Qwen-Image-2.1/blob/main/LICENSE
https://html.non.io/qwen-comparison/
The text rendering definitely is much, much better than anything else on the open weights market right now. Small text fidelity is quite good. It seems like the text encoder however gets a little bit overloaded with larger prompts - note the presence of hex codes in the design output, those were inputs from the expanded prompt.
I'll be trying a post-training run on this for web design, it has some serious potential.
[1] diffui.ai
None are as good yet, but what everyone said is coming true - models are not moats. And these folks need an exit (even Msuk whose shares are still locked)
(I mean: outside direct or substantial use of Python, and running the Neural Network in the most efficient way.)
In my experience, video models generate videos pretty well but are mid at editing. They actually regenerate the entire video along with the edit. So these models being non-deterministic tweak the rest of the video as well, the parts you hoped would be left not edited. It gets exponentially worse when there are humans in the videos, annoying face distortions and for some reason these models just don't understand fingers.
EDIT: It still produces artifacts it's better but unusable for production work. In midvalues you will see a slight dot pattern.
Excited to see what the future holds for them!
I know a few friends of mine who are running models and are ignoring the licence.
Whether it is AGPL 3.0, or a completely restrictive license, it is going to get broken anyway and be used for commercial purposes.
I don't know anyone who looks at the licenses of the OSS software they are using.
In today’s world OSS is synonymous with "Free" and the AI model providers are proof of that with their training of code, datasets, etc.
So it begs the question, why should we abide by their licenses of their models?
(Or does this model has those capabilities natively ?)
Code on github, models on huggingface, nice intro text: "We are excited to open-source Qwen-Image-2.1 [...]".
meh...