Tencent Hunyuan's Hy Image 3.5 preview lands in ComfyUI with +30% human eval win rate
TencentHunyuan · x · 2026-09-23
Tencent Hunyuan officially announced that Hy Image 3.5 preview is now available in ComfyUI, claiming a +30% win rate in human evaluation versus Hy Image 3.0.
Key capabilities:
- Text-to-image and image-to-image in one model, up to 2K resolution;
- Multilingual text, symbols, and small print rendered correctly;
- Styles spanning cinematic, comic, commercial photography, and illustration;
- Identity and product features that hold through scene, outfit, and style changes.
For designers and developers running ComfyUI production pipelines, this is a directly usable drop-in upgrade from a major Chinese lab.
More from Multimodal
- RULER: instance-aware rubric rewards beat reward hacking in SVG generation RL — inclusionAI · 2026-09-23
- Tencent ARC's GAE: geometry-native latent space cuts FVD by up to 23% in world generation — TencentARC · 2026-09-23
- PixVerse unveils R2 real-time world model: actions carry cause and effect — alifcoder · 2026-09-23
- Solo creator builds AI action short with MiniMax, Tripo and Blender — shares full workflow — Spoonman915 · 2026-09-23
- Game-to-CGI transition cuts with MiniMax H3 on an RTX 4060 — seed management tips included — Far_Cast_Far_Wide · 2026-09-23
- Pexo's conversational AI video agent churns out product ads for pennies — SimplyAnnisa · 2026-09-23