ComfyUI launches model usage rankings: Qwen Image 2.1 tops generations at 16%
cremebait15 · reddit · 2026-11-22
ComfyUI now publishes live model usage rankings at router.comfy.org, based on actual image generations (left) and share of users who ran each model at least once (right):
- Qwen Image 2.1 leads by generation share (16%) with 19% user coverage; Z-Image Turbo has the widest user reach (37%) at 8.6% of generations — possibly inflated as the default.
- Krea 2 Turbo overperforms: 10% of generations, used by 31% of users.
- Others: FLUX.2 Klein 9B 9.3% of generations but only 2.8% of users; MOSS Image 5.3%; Qwen Image Edit 2511 reaches 8.8% of users; FLUX.2 Dev just 1.6%.
The right column can exceed 100% since users try multiple models; accuracy is still unverified.
More from Multimodal
- pi-H3, a flow-matching video model built on MiniMax-H3, trends on Hugging Face — Lakonik · 2026-11-22
- GPT 6.2 Astra composes Beethoven-style scherzo, with all 604 key strikes from the score — imjustnewatai · 2026-11-22
- Musk demos real-time AI video generation on half of one AI4 TRIP accelerator — jamesdouma · 2026-11-22
- RAEA engine lets AI agents create videos, 3D scenes and games with React — Promptmethus · 2026-11-22
- Do adaptive optimizers still work? Reddit users doubt it for flow-matching models — Inevitable_Pen9043 · 2026-11-22
- SeaDream 5.5 image demo stuns with 2K photorealistic quality — gen_ericai · 2026-11-22