Qwen Image 2.1 🤩Viggle-turbo V0.2.1 -5× faster than the 40-step base model end to end, and very competitive with it in quality . -The clearest gap is small, dense text -It does both text-to-image and instruction-driven editing with 1–3 reference images

Qwen-Image-2.1-viggle-turbo v0.2 is out on Hugging Face Text-to-image and image editing in 6 steps about 5× faster than the 40-step Qwen-Image-2.1 model:

Qwen Image 2.1 💡🙂☀️Natural-Exposure-LoRA Designed to transform images with balanced, neutral exposure while preserving natural colors, details, lighting, and overall image consistency.

Qwen’s Image 2.1 released but it come with lots of improvements but just as many caveats. The old Qwen-Image-2512 was a 20B model with a massive: 💾 40.9GB BF16 image transformer But the new Qwen-Image-2.1 comes with … 🧠 7B visual generator 💾 14.2GB BF16 🔥 7.26GB INT8 already available for ComfyUI 👈👀 (Day-1) This little thing can … 🎨 generate AND edit images 🖼️ generate native 2K 🫥 create real transparent RGBA images 👥 use up to 10 reference images ✏️ preserve people/products while editing 🔤 render text 🎯 do masked/local edits Qwen basically took the huge local image model and reduced with a catch … ⚠️ The tiny 7.26GB image model comes with an entire pipeline. It uses a Qwen3-VL 8B encoder + VAE. ComfyUI already has a 6.31GB W4A8 encoder, and Qwen supports CPU offloading for smaller GPUs. Also… 🚫 Qwen-Image-2.1 is NON-COMMERCIAL under the new Qwen Research License. That’s especially notable because Qwen-Image-2512 was Apache 2.0.

Qwen Image 3.0 is now on OpenArt ✨ The most Real Qwen image model yet. Native text across 12 languages, precise 10px type, and full interfaces like web pages, games, and livestreams rendered with real world knowledge. Available now on OpenArt:

Qwen Image 3 Pro by @Alibaba_Qwen is 8th overall on Image Arena with an Elo of 1282. This is a 25 rank and 89 Elo jump from their previous model, Wan 2.7 Pro. Qwen Image 3 Pro is particularly strong at Graphic Design, ranked 5th ahead of Seedream 5.0 Pro by @BytePlusGlobal. This establishes @Alibaba_Qwen among the top 5 labs overall on Image Arena, following @GoogleDeepMind and @BytePlusGlobal. Congratulations to the @Alibaba_Qwen team on the accomplishment!

Qwen‑Image‑3.0 is built for complex content creation, from detailed knowledge diagrams to UI interfaces, all in one step. With ultra‑long inputs of up to 4.5K tokens and native rendering in 12 languages, Qwen‑Image‑3.0 unlocks new possibilities for creators and developers alike. 🎨Discover what you can build next!

Qwen-Image-2.0-RL Technical Report Paper: They fine-tune VLMs to score pointwise with CoT, so the reward judges alignment against an explanation and is harder to game. OPD then closes the train-inference gap on the diffusion backbone.