• 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.
[1] diffui.ai
show comments
fishfasell
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.
show comments
mdp2021
How do you use this model locally, similarly to using `llama-server -m <model>`?
(I mean: outside direct or substantial use of Python, and running the Neural Network in the most efficient way.)
show comments
trains39472
A 7B diffusion model can now render CJK text better than Microsoft Windows.
show comments
thenipper
The uncanny valley cheers is really freaking me out.
d2kx
God I love the Qwen team. Easily the most diverse set of models from all the Chinese labs. Only Gemini/DeepMind comes close.
trentor
They finally fixed their VAE. It really held back their models over the last 2 years.
EDIT: It still produces artifacts it's better but unusable for production work. In midvalues you will see a slight dot pattern.
show comments
gunalx
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.
show comments
docheinestages
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.
show comments
hgufj
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.
show comments
Hard_Space
Interesting in the example of assembling the Cheers team how the otherwise great result genericizes Shelley Long.
show comments
samayashar
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.
Excited to see what the future holds for them!
amelius
I don't want only cherry picked examples. Show me failure modes too.
show comments
TomGarden
Very impressive, and kind of worrying a 7B model can have such capabilities. The implications are huge. And Qwen does no watermarking (yet) yeah?
show comments
spottedmarley
Boy do I love waking up to find a new awesome toy from the Qwen team waiting for me to play with! Pulling it now
colesantiago
While the license of this model is a shame it is still unenforceable.
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?
show comments
timmytokyo
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.
finnjohnsen2
I would call this a license trap: Qwen RESEARCH LICENSE AGREEMENT
Code on github, models on huggingface, nice intro text: "We are excited to open-source Qwen-Image-2.1 [...]".
meh...
hirako2000
Woody Harrelson?
ramesh31
Still fails to generate smoothly animated sprites, although the native RGBA transparency is nice. Anyone found one that can?
jonplackett
Is it just me or are Alibaba / qwen’s websites often appear broken / very slow?
bknight1983
While I'm impressed with the Bluey example, the lack of Muffin disappoints me.
hn45e7pbij
Image gen you eyeball one frame and stop, code needs hundreds of tokens all correct in sequence, one bad line and the whole thing fails.
So thoughts
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
A lot of the previous Qwen models seem to have used Apache licenses, among others:
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
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:
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
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.
How do you use this model locally, similarly to using `llama-server -m <model>`?
(I mean: outside direct or substantial use of Python, and running the Neural Network in the most efficient way.)
A 7B diffusion model can now render CJK text better than Microsoft Windows.
The uncanny valley cheers is really freaking me out.
God I love the Qwen team. Easily the most diverse set of models from all the Chinese labs. Only Gemini/DeepMind comes close.
They finally fixed their VAE. It really held back their models over the last 2 years.
EDIT: It still produces artifacts it's better but unusable for production work. In midvalues you will see a slight dot pattern.
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.
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.
Interesting in the example of assembling the Cheers team how the otherwise great result genericizes Shelley Long.
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.
Excited to see what the future holds for them!
I don't want only cherry picked examples. Show me failure modes too.
Very impressive, and kind of worrying a 7B model can have such capabilities. The implications are huge. And Qwen does no watermarking (yet) yeah?
Boy do I love waking up to find a new awesome toy from the Qwen team waiting for me to play with! Pulling it now
While the license of this model is a shame it is still unenforceable.
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?
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.
I would call this a license trap: Qwen RESEARCH LICENSE AGREEMENT
Code on github, models on huggingface, nice intro text: "We are excited to open-source Qwen-Image-2.1 [...]".
meh...
Woody Harrelson?
Still fails to generate smoothly animated sprites, although the native RGBA transparency is nice. Anyone found one that can?
Is it just me or are Alibaba / qwen’s websites often appear broken / very slow?
While I'm impressed with the Bluey example, the lack of Muffin disappoints me.
Image gen you eyeball one frame and stop, code needs hundreds of tokens all correct in sequence, one bad line and the whole thing fails.