Qwen-Image

“I found that for Qwen Image 2.1 the prompt enhancer makes a big difference.”

“uncensored Qwen-Image then tell it where the ads are and what to replace them with”

“Flux 2 Klein or Qwen Image 2.1 might give you a less frustrating experience at comparable or better performance.”

“It's bafflingly worse than other comparably sized ones: Qwen-Image 2.1 7B”

“a rather impressive example of what sounds just like this in Qwen recently”

“Qwen-Image 2.1 is definitely a pretty big leap over the last open-weight version”

“A lot of people were putting ZiT as a refiner downstream in early Qwen-Image 1.0 workflows”

“Qwen is trained off of gpt's outputs. This is both a positive and negative”

“Still fails to generate smoothly animated sprites, although the native RGBA transparency is nice.”

“the license seems to cover any commercial usage at all, which would certainly include self-hosting for internal business purposes”

“it seems a lot better at understanding and following Chinese prompts”

“It's fast using QwenImage2.1 convrot, a 1MP image took around ~5 seconds on an RTX4090.”

“The text rendering definitely is much, much better than anything else on the open weights market right now.”

“What this research license achieves is that we cannot use this model in applications we publish.”

“I am disappointed that the model is not actually open-weights as I expected based on the headline.”

“now with the restriction on commercial use (which is against opensource definition) it is not even open-weights”

“I love the non-commercial clauses because of how many people are using these for deceptive ads”

“While I'm impressed with the Bluey example, the lack of Muffin disappoints me.”

“You shall not use the Materials for any commercial purpose without obtaining a separate commercial license from us.”

“It still produces artifacts it's better but unusable for production work.”

“A 7B diffusion model can now render CJK text better than Microsoft Windows.”

“Boy do I love waking up to find a new awesome toy from the Qwen team”

“Easily the most diverse set of models from all the Chinese labs.”

“Very impressive, and kind of worrying a 7B model can have such capabilities.”

“It's a big step up over their prior model, but it's nowhere near gpt-image-2”

“Everything is less good than gpt-2-image and I suspect that will be the case for awhile”

“it's ability to things like render images that look like they were rendered with LaTeX”

“It's not authoritative, by any means, but it did answer some questions”

“From the images I've seen, Qwen-Image 3.0 is just a subpar equivalent to other proprietary models like gpt-image-2 and nb-pro.”

“their woman face generation always use very similar, too perfect, same prettiness faces”

“The examples posted on their launch blog page are quite impressive, especially for fine details”

“effortlessly generating complex layouts such as newspapers, storyboards, and exam papers”

“there's been a lot of positive movement in the open-weights space”

“Qwen-Image-2512 / Z-Image / Flux.2 absolutely crush SDXL if you're actually generating moderately complex scenes.”

“Qwen was already excellent and now they rolled Image and Edit together for an “Omni” model.”

“The complex prompt following ability and editing is seriously impressive here.”

“it's worse at all of those things other than perhaps looking like it was shot on a cell phone camera than Qwen Image and Flux 2”

“significantly better prompt understanding but has an unnatural "smoothness" to its generated images”

“has significantly better prompt understanding but has an unnatural "smoothness" to its generated images”

“the Chinese models are easily the best or very close to the best”