{"slug": "qwen-image", "name": "Qwen-Image", "category": "ai-models", "category_name": "AI models", "description": "", "recommenders": 7, "commenters": 13, "warners": 4, "score": 2.652, "rising": 4.02, "first": "2025-11-20", "latest": "2026-10-25", "why": "With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera.", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 8, "neutral": 5, "compare": 1, "warn": 5}, "spark": [1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 3, 1], "listed": true, "loved": [], "caveat": "", "short_quote": "With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera.", "kind": "", "icon": true, "family": "qwen", "domain": null, "github_repo": "github.com/qwenlm/qwen-image-2.1", "official_url": "https://github.com/QwenLM/Qwen-Image", "pricing": "", "aliases": ["Qwen Image 2.1", "Qwen-Image", "Qwen-Image-2512", "Qwen-Image-2.0", "Qwen-Image 2.1"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 803782177, "item_id": 50236386, "url": "https://news.ycombinator.com/item?id=50236386", "by": "razster", "date": "2026-10-25", "version": "2.1", "stance": "recommend", "quote": "most are done with Qwen Image 2.1", "excerpt": "Here are some of my prompts. https://old.reddit.com/r/StableDiffusion/comments/1xb59yw/fibo_2_edit_is_what_klein_3_shouldve_been/19rgnauzvf85g/ I use the Portable version of ComfyUI, most are done with Qwen Image 2.1 and some hard restores using FIBO2. You can find a bunch of video tutorials on YT for both models.", "context": "", "use_case": "image editing", "switched_from": "", "story": "Nikon photography contest winner loses title after scientists spot signs of AI"}, {"id": 803769666, "item_id": 50235604, "url": "https://news.ycombinator.com/item?id=50235604", "by": "razster", "date": "2026-10-25", "version": "2.1", "stance": "recommend", "quote": "once I started using ComfyUI with FIBO2 and Qwen Image 2.1 to restore my photographs, I was hook!", "excerpt": "Oh yeah, I'm not going to lie, once I started using ComfyUI with FIBO2 and Qwen Image 2.1 to restore my photographs, I was hook! I have 25+ years of digital and film photos that I've been restoring due to my older camera/lens quality. The addiction of restoring these older photos and then processing them once again in either Lightroom or Photoshop. I caught myself spending countless hours revisiting the past and rest", "context": "hobby photographer", "use_case": "restoring photographs", "switched_from": "", "story": "Nikon photography contest winner loses title after scientists spot signs of AI"}, {"id": 803767041, "item_id": 50235440, "url": "https://news.ycombinator.com/item?id=50235440", "by": "razster", "date": "2026-10-25", "version": "2.1", "stance": "recommend", "quote": "With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera.", "excerpt": "Oh man, I love using AI to retouch my photographs. I shoot with a Canon 5D Mark I (OG), which has a faulty sensor. Since I can\u2019t afford an upgrade, I rely on vintage glass: Pentax SMC 50mm and 35mm lenses adapted from M42 to EOS. With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera. I then pass them through Lightroom for fine adjustments,", "context": "hobby photographer, 25 years", "use_case": "retouching photographs", "switched_from": "", "story": "Nikon photography contest winner loses title after scientists spot signs of AI"}, {"id": 796483440, "item_id": 49780215, "url": "https://news.ycombinator.com/item?id=49780215", "by": "vunderba", "date": "2026-09-20", "version": "2.1", "stance": "recommend", "quote": "Qwen-Image 2.1 is definitely a pretty big leap over the last open-weight version", "excerpt": "Well, the results are in, at least for text-to-image (the editing bench will come later). Qwen-Image 2.1 is definitely a pretty big leap over the last open-weight version, Qwen-Image 1.0, released back in August of last year and managed to score 7 out of 15 as opposed to its predecessor which scored 4 out of 15. Even though it's significantly smaller, 7b vs 20b, it's multimodal (so you don't need a separate image-to-", "context": "runs image model benchmarks", "use_case": "text-to-image generation", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796439121, "item_id": 49777445, "url": "https://news.ycombinator.com/item?id=49777445", "by": "peri-cl", "date": "2026-09-20", "version": "2.1", "stance": "recommend", "quote": "very easy", "excerpt": "I've set it up on my local machine just now, as my first local image diffuser. I can confirm it's very easy. I tried stable-diffusion.cpp, following its compile guide here[0], and its Qwen Image-2.1 specific instructions here[1]. It works out of the box. I made a test pelican[2]. It took 3 minutes on a CPU. [0] https://github.com/leejet/stable-diffusion.cpp/blob/master/docs/build.md [1] https://github.com/leejet/stab", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 802034945, "item_id": 50127184, "url": "https://news.ycombinator.com/item?id=50127184", "by": "vunderba", "date": "2026-10-17", "version": null, "stance": "warn", "quote": "the original Qwen-Image was actually Apache 2 before they switched over", "excerpt": "Yeah, unfortunately, a lot of them have been moving to their own bespoke, non-commercial licenses (Ideogram, Krea, Flux, Qwen 2, etc). Annoyingly, the original Qwen-Image was actually Apache 2 before they switched over unfortunately. In addition to Moss, Boogu Image 0.1 is a unified model which came out in June and is kind of being slept on right now, is actually pretty decent and falls under the Apache 2.0 license. ", "context": "", "use_case": "", "switched_from": "", "story": "FLUX 3 Image Klein"}, {"id": 796419968, "item_id": 49776248, "url": "https://news.ycombinator.com/item?id=49776248", "by": "jfoster", "date": "2026-09-20", "version": "2.1", "stance": "warn", "quote": "You shall not use the Materials for any commercial purpose without obtaining a separate commercial license from us.", "excerpt": "Note that the license on this has this in it: > You shall not use the Materials for any commercial purpose without obtaining a separate commercial license from us. It probably will be much cheaper to use than other image models, but it seems that will be up to the whims of Qwen/Alibaba rather than just being the cost of putting it in a cloud provider. https://github.com/QwenLM/Qwen-Image-2.1/blob/main/LICENSE", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796418160, "item_id": 49776135, "url": "https://news.ycombinator.com/item?id=49776135", "by": "jfoster", "date": "2026-09-20", "version": "2.1", "stance": "warn", "quote": "it looks like this model is using a much more restrictive license", "excerpt": "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", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 771201427, "item_id": 48200089, "url": "https://news.ycombinator.com/item?id=48200089", "by": "vunderba", "date": "2026-05-19", "version": null, "stance": "warn", "quote": "even the state-of-the-art local models are probably only about thirty percent as good", "excerpt": "Definitely. I run an entire site built around a series of benchmarks that focus on prompts of increasingly difficult complexity with a focus on adherence, and even the state-of-the-art local models are probably only about thirty percent as good as proprietary models like Gemini 3.1 Flash Image and GPT Image 2. Comparing Qwen-Image, Flux.2, ZiT, NB2, and gpt-image-2 https://genai-showdown.specr.net/?models=qi,nbp3,f2d", "context": "", "use_case": "", "switched_from": "", "story": "OpenAI Adopts Google's SynthID Watermark for AI Images with Verification Tool"}]}], "who": [{"by": "razster", "context": "hobby photographer", "url": "https://news.ycombinator.com/item?id=50235604", "date": "2026-10-25"}, {"by": "vunderba", "context": "runs image model benchmarks", "url": "https://news.ycombinator.com/item?id=49780215", "date": "2026-09-20"}], "tags": [], "use_cases": ["image generation", "text-to-image generation", "image editing", "restoring photographs", "retouching photographs", "comic panel image generation"], "compared": [{"slug": "z-image-turbo", "name": "Z-Image Turbo", "co": 5, "switched_from": 0, "switched_to": 0}, {"slug": "flux", "name": "FLUX", "co": 4, "switched_from": 0, "switched_to": 0}, {"slug": "comfyui", "name": "ComfyUI", "co": 3, "switched_from": 0, "switched_to": 0}, {"slug": "krea-2", "name": "Krea 2", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "adobe-lightroom", "name": "Adobe Lightroom", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "gpt-image", "name": "GPT Image", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49775499", "title": "Qwen Image 2.1", "mentions": 7}, {"url": "https://news.ycombinator.com/item?id=46957198", "title": "Qwen-Image-2.0: Professional infographics, exquisite photorealism", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=50229827", "title": "Nikon photography contest winner loses title after scientists spot signs of AI", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=50123580", "title": "FLUX 3 Image Klein", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49819252", "title": "Qwen Image 2.1 beats Google Nano Banana 2.0 with minuscule 7B parameter model", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49023016", "title": "Startup founders urge Trump not to shut off Chinese open weight AI", "mentions": 1}], "sentiment_by_month": {"2025-11": {"recommend": 1}, "2025-12": {"neutral": 2, "recommend": 1}, "2026-01": {"neutral": 1, "warn": 1, "compare": 1}, "2026-02": {"neutral": 1, "recommend": 1, "warn": 1}, "2026-03": {}, "2026-04": {}, "2026-05": {"warn": 1}, "2026-06": {}, "2026-07": {"recommend": 1}, "2026-08": {}, "2026-09": {"recommend": 3, "warn": 1, "neutral": 1}, "2026-10": {"warn": 1, "recommend": 1}}, "versions": [], "version_unnamed": 7, "family_ref": {"slug": "qwen", "name": "Qwen"}, "runs_on": null, "lines": [], "quotes": [{"id": 803782177, "item_id": 50236386, "url": "https://news.ycombinator.com/item?id=50236386", "by": "razster", "date": "2026-10-25", "version": "2.1", "stance": "recommend", "quote": "most are done with Qwen Image 2.1", "excerpt": "Here are some of my prompts. https://old.reddit.com/r/StableDiffusion/comments/1xb59yw/fibo_2_edit_is_what_klein_3_shouldve_been/19rgnauzvf85g/ I use the Portable version of ComfyUI, most are done with Qwen Image 2.1 and some hard restores using FIBO2. You can find a bunch of video tutorials on YT for both models.", "context": "", "use_case": "image editing", "switched_from": "", "story": "Nikon photography contest winner loses title after scientists spot signs of AI"}, {"id": 803769666, "item_id": 50235604, "url": "https://news.ycombinator.com/item?id=50235604", "by": "razster", "date": "2026-10-25", "version": "2.1", "stance": "recommend", "quote": "once I started using ComfyUI with FIBO2 and Qwen Image 2.1 to restore my photographs, I was hook!", "excerpt": "Oh yeah, I'm not going to lie, once I started using ComfyUI with FIBO2 and Qwen Image 2.1 to restore my photographs, I was hook! I have 25+ years of digital and film photos that I've been restoring due to my older camera/lens quality. The addiction of restoring these older photos and then processing them once again in either Lightroom or Photoshop. I caught myself spending countless hours revisiting the past and rest", "context": "hobby photographer", "use_case": "restoring photographs", "switched_from": "", "story": "Nikon photography contest winner loses title after scientists spot signs of AI"}, {"id": 803767041, "item_id": 50235440, "url": "https://news.ycombinator.com/item?id=50235440", "by": "razster", "date": "2026-10-25", "version": "2.1", "stance": "recommend", "quote": "With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera.", "excerpt": "Oh man, I love using AI to retouch my photographs. I shoot with a Canon 5D Mark I (OG), which has a faulty sensor. Since I can\u2019t afford an upgrade, I rely on vintage glass: Pentax SMC 50mm and 35mm lenses adapted from M42 to EOS. With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera. I then pass them through Lightroom for fine adjustments,", "context": "hobby photographer, 25 years", "use_case": "retouching photographs", "switched_from": "", "story": "Nikon photography contest winner loses title after scientists spot signs of AI"}, {"id": 802034945, "item_id": 50127184, "url": "https://news.ycombinator.com/item?id=50127184", "by": "vunderba", "date": "2026-10-17", "version": null, "stance": "warn", "quote": "the original Qwen-Image was actually Apache 2 before they switched over", "excerpt": "Yeah, unfortunately, a lot of them have been moving to their own bespoke, non-commercial licenses (Ideogram, Krea, Flux, Qwen 2, etc). Annoyingly, the original Qwen-Image was actually Apache 2 before they switched over unfortunately. In addition to Moss, Boogu Image 0.1 is a unified model which came out in June and is kind of being slept on right now, is actually pretty decent and falls under the Apache 2.0 license. ", "context": "", "use_case": "", "switched_from": "", "story": "FLUX 3 Image Klein"}, {"id": 797110640, "item_id": 49819415, "url": "https://news.ycombinator.com/item?id=49819415", "by": "vunderba", "date": "2026-09-23", "version": "2.1", "stance": "neutral", "quote": "the quality is frustratingly inconsistent", "excerpt": "I haven't done enough tests on the editing capabilities but as far as the strict text to image goes, Alibaba is reaching to claim it is comparable to NB 2. It's definitely a nice upgrade from the last open-weight version (Qwen-Image 1.0) at least in terms of strict coherence and prompt understanding... but a lot of the outputs seem distilled for lack of a better word - likely trained on poor synthetic data. There's a", "context": "", "use_case": "Text-to-image generation", "switched_from": "", "story": "Qwen Image 2.1 beats Google Nano Banana 2.0 with minuscule 7B parameter model"}, {"id": 796483440, "item_id": 49780215, "url": "https://news.ycombinator.com/item?id=49780215", "by": "vunderba", "date": "2026-09-20", "version": "2.1", "stance": "recommend", "quote": "Qwen-Image 2.1 is definitely a pretty big leap over the last open-weight version", "excerpt": "Well, the results are in, at least for text-to-image (the editing bench will come later). Qwen-Image 2.1 is definitely a pretty big leap over the last open-weight version, Qwen-Image 1.0, released back in August of last year and managed to score 7 out of 15 as opposed to its predecessor which scored 4 out of 15. Even though it's significantly smaller, 7b vs 20b, it's multimodal (so you don't need a separate image-to-", "context": "runs image model benchmarks", "use_case": "text-to-image generation", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796439121, "item_id": 49777445, "url": "https://news.ycombinator.com/item?id=49777445", "by": "peri-cl", "date": "2026-09-20", "version": "2.1", "stance": "recommend", "quote": "very easy", "excerpt": "I've set it up on my local machine just now, as my first local image diffuser. I can confirm it's very easy. I tried stable-diffusion.cpp, following its compile guide here[0], and its Qwen Image-2.1 specific instructions here[1]. It works out of the box. I made a test pelican[2]. It took 3 minutes on a CPU. [0] https://github.com/leejet/stable-diffusion.cpp/blob/master/docs/build.md [1] https://github.com/leejet/stab", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796431632, "item_id": 49776977, "url": "https://news.ycombinator.com/item?id=49776977", "by": "vunderba", "date": "2026-09-20", "version": "2.1", "stance": "neutral", "quote": "this one explicitly forbids commercial usage without obtaining a separate license", "excerpt": "So thoughts Positives \u2022 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. \u2022 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 back", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796430352, "item_id": 49776897, "url": "https://news.ycombinator.com/item?id=49776897", "by": "jjcm", "date": "2026-09-20", "version": "2.1", "stance": "recommend", "quote": "The text rendering definitely is much, much better than anything else on the open weights market right now.", "excerpt": "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", "context": "", "use_case": "Image generation", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796419968, "item_id": 49776248, "url": "https://news.ycombinator.com/item?id=49776248", "by": "jfoster", "date": "2026-09-20", "version": "2.1", "stance": "warn", "quote": "You shall not use the Materials for any commercial purpose without obtaining a separate commercial license from us.", "excerpt": "Note that the license on this has this in it: > You shall not use the Materials for any commercial purpose without obtaining a separate commercial license from us. It probably will be much cheaper to use than other image models, but it seems that will be up to the whims of Qwen/Alibaba rather than just being the cost of putting it in a cloud provider. https://github.com/QwenLM/Qwen-Image-2.1/blob/main/LICENSE", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796418160, "item_id": 49776135, "url": "https://news.ycombinator.com/item?id=49776135", "by": "jfoster", "date": "2026-09-20", "version": "2.1", "stance": "warn", "quote": "it looks like this model is using a much more restrictive license", "excerpt": "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", "context": "", "use_case": "", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 796415552, "item_id": 49775972, "url": "https://news.ycombinator.com/item?id=49775972", "by": "spottedmarley", "date": "2026-09-20", "version": "2.1", "stance": "recommend", "quote": "Boy do I love waking up to find a new awesome toy from the Qwen team", "excerpt": "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", "context": "", "use_case": "image generation", "switched_from": "", "story": "Qwen Image 2.1"}, {"id": 784383969, "item_id": 49023998, "url": "https://news.ycombinator.com/item?id=49023998", "by": "vunderba", "date": "2026-07-23", "version": null, "stance": "recommend", "quote": "Qwen-Image-2512/Qwen-Edit-2511. They are both decent", "excerpt": "The two major image generation models are both by Alibaba - Z-Image and Qwen-Image-2512/Qwen-Edit-2511. They are both decent but have different use-cases. Unfortunately with the recent Qwen-Image 2 and now 3 it doesn't seem likely that they're going to be doing any more open-weight image gen. https://huggingface.co/Tongyi-MAI/Z-Image-Turbo https://huggingface.co/Qwen/Qwen-Image-2512 https://huggingface.co/Qwen/Qwen-I", "context": "", "use_case": "", "switched_from": "", "story": "Startup founders urge Trump not to shut off Chinese open weight AI"}, {"id": 771201427, "item_id": 48200089, "url": "https://news.ycombinator.com/item?id=48200089", "by": "vunderba", "date": "2026-05-19", "version": null, "stance": "warn", "quote": "even the state-of-the-art local models are probably only about thirty percent as good", "excerpt": "Definitely. I run an entire site built around a series of benchmarks that focus on prompts of increasingly difficult complexity with a focus on adherence, and even the state-of-the-art local models are probably only about thirty percent as good as proprietary models like Gemini 3.1 Flash Image and GPT Image 2. Comparing Qwen-Image, Flux.2, ZiT, NB2, and gpt-image-2 https://genai-showdown.specr.net/?models=qi,nbp3,f2d", "context": "", "use_case": "", "switched_from": "", "story": "OpenAI Adopts Google's SynthID Watermark for AI Images with Verification Tool"}, {"id": 751427728, "item_id": 46964233, "url": "https://news.ycombinator.com/item?id=46964233", "by": "thisisit", "date": "2026-02-10", "version": "2", "stance": "warn", "quote": "so, not very reliable in my books.", "excerpt": "I liked their comic panels example and tried it using their chat at: https://chat.qwen.ai/ When I used the exact prompt the post - the chat works. It gives me the exact output from the blog post. Then I used Google Translate to understand the prompt format. The prompt is: A 4x6 panel comic, four lines, six panels per line. Each panel is separated by a white dividing line. The first row, from left to right: Panel 1: P", "context": "", "use_case": "comic panel image generation", "switched_from": "", "story": "Qwen-Image-2.0: Professional infographics, exquisite photorealism"}, {"id": 751382688, "item_id": 46961418, "url": "https://news.ycombinator.com/item?id=46961418", "by": "vunderba", "date": "2026-02-10", "version": null, "stance": "recommend", "quote": "The original Qwen-Image scored the highest among local models for image editing", "excerpt": "Couple of thoughts: 1. I\u2019d wager that given their previous release history, this will be open\u2011weight within 3-4 weeks. 2. It looks like they\u2019re following suit with other models like Z-Image Turbo (6B parameters) and Flux.2 Klein (9B parameters), aiming to release models that can run on much more modest GPUs. For reference, the original Qwen-Image is a 20B-parameter model. 3. This is a unified model (both image genera", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-Image-2.0: Professional infographics, exquisite photorealism"}, {"id": 751382689, "item_id": 46961418, "url": "https://news.ycombinator.com/item?id=46961418", "by": "vunderba", "date": "2026-02-10", "version": "2", "stance": "neutral", "quote": "I'll probably be waiting until the local version drops", "excerpt": "Couple of thoughts: 1. I\u2019d wager that given their previous release history, this will be open\u2011weight within 3-4 weeks. 2. It looks like they\u2019re following suit with other models like Z-Image Turbo (6B parameters) and Flux.2 Klein (9B parameters), aiming to release models that can run on much more modest GPUs. For reference, the original Qwen-Image is a 20B-parameter model. 3. This is a unified model (both image genera", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-Image-2.0: Professional infographics, exquisite photorealism"}, {"id": 751323649, "item_id": 46957728, "url": "https://news.ycombinator.com/item?id=46957728", "by": "Mashimo", "date": "2026-02-10", "version": null, "stance": "neutral", "quote": "", "excerpt": "I had no problems getting images with blurry background with the appropriate prompts. Something like \"shallow depth of fields, bokeh, DSLR\" can lead to good results. https://cdn.discordapp.com/attachments/1180506623475720222/1456789321091649769/image.png?ex=698bbd47&is=698a6bc7&hm=d949f70e4ac8b0173f8addc6947a1bade5482b54e592712da1c1950166f30f42& [0] But I found that that results in more professional looking images, a", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-Image-2.0: Professional infographics, exquisite photorealism"}, {"id": 748795200, "item_id": 46799700, "url": "https://news.ycombinator.com/item?id=46799700", "by": "vunderba", "date": "2026-01-28", "version": null, "stance": "compare", "quote": "local (Qwen-Image, Flux) vs proprietary (NB Pro, Seedream)", "excerpt": "Cool - GenAI image generation is a deep rabbit hole that you're about to fall into! Super happy that you pit LLMs against relatively recent IF to mitigate cheating through pre-existing training data as well. FYI I've been running a SOTA model comparison site for about a year now that looks at prompt adherence across local (Qwen-Image, Flux) vs proprietary (NB Pro, Seedream) that might help give an idea where the capa", "context": "", "use_case": "", "switched_from": "", "story": "Updated LLM Benchmark (Gemini 3 Flash)"}, {"id": 746492578, "item_id": 46655786, "url": "https://news.ycombinator.com/item?id=46655786", "by": "CamperBob2", "date": "2026-01-16", "version": null, "stance": "warn", "quote": "Qwen-Image is just plain not as good", "excerpt": "Not in my experience. Flux 2 is much larger and heavily censored, and Qwen-Image is just plain not as good. You can fool me into thinking that Z-Image Turbo output isn't AI, while that's rarely the case with Qwen. Look at the images I posted elsewhere in this section. They are crappy excuses for pogo sticks, but they absolutely do NOT look like they came from a cell phone. Also see vunderba's page at https://genai-sh", "context": "", "use_case": "image generation", "switched_from": "", "story": "FLUX.2 [Klein]: Towards Interactive Visual Intelligence"}, {"id": 743648481, "item_id": 46478030, "url": "https://news.ycombinator.com/item?id=46478030", "by": "yanaimngvov", "date": "2026-01-03", "version": null, "stance": "neutral", "quote": "If you want to test Alibaba's new Qwen-Image-2512 model", "excerpt": "If you want to test Alibaba's new Qwen-Image-2512 model but don't have the GPU memory (or patience) to host it locally, we put up a free instance here: https://zimage.run It currently supports Qwen-Image-2512 and Z-Image Turbo. No account needed. It might be a bit slow if traffic spikes, but it works for quick prompt testing.", "context": "", "use_case": "", "switched_from": "", "story": "A free, no-login Web UI for Qwen-Image-2512"}, {"id": 743111952, "item_id": 46444497, "url": "https://news.ycombinator.com/item?id=46444497", "by": "homarp", "date": "2025-12-31", "version": null, "stance": "neutral", "quote": "Qwen released their updated Image model", "excerpt": "Qwen released their updated Image model: https://huggingface.co/Qwen/Qwen-Image-2512 One way to run it is with ComfyUI https://www.comfy.org/ Or you can try https://github.com/leejet/stable-diffusion.cpp", "context": "", "use_case": "", "switched_from": "", "story": "How to Run Qwen-Image-2512 Locally in ComfyUI"}, {"id": 743086752, "item_id": 46442922, "url": "https://news.ycombinator.com/item?id=46442922", "by": "danielhanchen", "date": "2025-12-31", "version": null, "stance": "recommend", "quote": "Qwen's latest Qwen-Image-2512 is currently the strongest open source model.", "excerpt": "Qwen's latest Qwen-Image-2512 is currently the strongest open source model. To run them locally, we made some GGUFs: https://huggingface.co/unsloth/Qwen-Image-2512-GGUF", "context": "", "use_case": "image generation", "switched_from": "", "story": "Qwen-Image-2512"}, {"id": 738801089, "item_id": 46175068, "url": "https://news.ycombinator.com/item?id=46175068", "by": "vunderba", "date": "2025-12-06", "version": null, "stance": "neutral", "quote": "which has significantly better prompt understanding but has an unnatural \"smoothness\"", "excerpt": "I've done some preliminary testing with Z-Image Turbo in the past week. Thoughts - It's fast (~3 seconds on my RTX 4090) - Surprisingly capable of maintaining image integrity even at high resolutions (1536x1024, sometimes 2048x2048) - The adherence is impressive for a 6B parameter model Some tests (2 / 4 passed): https://imgpb.com/exMoQ Personally I find it works better as a refiner model downstream of Qwen-Image 20b", "context": "", "use_case": "", "switched_from": "", "story": "Z-Image: Powerful and highly efficient image generation model with 6B parameters"}, {"id": 735960352, "item_id": 45997522, "url": "https://news.ycombinator.com/item?id=45997522", "by": "tezza", "date": "2025-11-20", "version": null, "stance": "recommend", "quote": "Qwen won the latest models round last month", "excerpt": "Qwen won the latest models round last month\u2026 https://generative-ai.review/2025/09/september-2025-image-generation-qwen-image-vs-google-imagegen-v4-vs-google-nano-banana/ (non-pro Nano Banana)", "context": "", "use_case": "", "switched_from": "", "story": "Nano Banana Pro"}]}