signalAI热榜2026-10-02
Artificial Analysis Evaluation: Qwen-Image-2.1 Tops Both AA-Image Leaderboards as the Top Open-Weight Model
Qwen-Image-2.1, released by Alibaba on September 20 as an open-weight model, ranks #18 on both AA-Image-T2I v2.0 and AA-Image-Editing v2.0, placing it first among all open-weight models and ahead of Ideogram 4.0 (Quality) and HunyuanImage 3.0 Instruct. The 7B-parameter model handles text-to-image and image editing in one model, with native 2K output and RGBA support. Compared to Qwen Image 2.0, it jumped from #72 to #18 on T2I and from #58 to #18 on editing.
- for who
- AI developers and researchers evaluating open-weight image models
- why now
- Qwen-Image-2.1, released September 20, tops open-weight image rankings, making it worth evaluating now.
- what changes
- Open-weight image generation now has a top-tier option that rivals proprietary models on both generation and editing benchmarks
- to do
- Download the Qwen-Image-2.1 weights from Hugging Face or ModelScope and try it via Diffusers, ComfyUI, or vLLM-Omni
key points
- Qwen-Image-2.1 ranks #18 on both AA-Image leaderboards, first among open-weight models
- 7B-parameter model with native 2K output and RGBA support
- Outperforms Ideogram 4.0 and HunyuanImage 3.0; available on Hugging Face and ModelScope
#open source model#image generation#evaluation
score
score 8 out of 10. 0-10: how dense the facts are, multiplied by how much you can do with them after reading. 8+ means the topic's evidence bar is met: benchmarks and availability for a new model, amount and investors for a funding round, revenue figures for a solo-money story. Below 5 an item does not enter the digest. A press release scores 3 or less, a reprint loses 2, anything older than 14 days loses 1, a headline that misleads loses 3.
read the source