Starstruk Logo MarkStarstruk

v0.3.60

Comfy-Org/ComfyUIv0.3.60Sep 23, 2025by github-actions[bot]

AI Summary

This release introduces extensive support for the Hunyuan image model, including distilled and refiner variants, along with V3 schema conversions for various API nodes like Kling, Wan, and Cosmos. It also adds fp8 ops support for specific hardware and updates the frontend.

Key Highlights

  • Support for Hunyuan image distilled model and refiner with noise augmentation
  • V3 schema conversion for Kling, Wan, Moonvalley, and Cosmos API nodes
  • Added fp8 ops support by default on gfx1200 hardware
  • New WanAnimateToVideo node extensions for easier video generation
  • Support for qwen edit plus model and TextEncodeQwenImageEditPlus node

New Features

  • Hunyuan image model support
  • Runtime selection of attention functions
  • API nodes: Wan T2I, T2V, I2V nodes
  • Support for Chroma Radiance
  • Support for omnigen2 umo lora

Full Release Notes

## What's Changed
* Support hunyuan image distilled model. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9807
* Update template to 0.1.81 by @comfyui-wiki in https://github.com/comfyanonymous/ComfyUI/pull/9811
* Fast preview for hunyuan image. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9814
* Implement hunyuan image refiner model. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9817
* Add Output to V3 Combo type to match what is possible with V1 by @Kosinkadink in https://github.com/comfyanonymous/ComfyUI/pull/9813
* Bump frontend to 1.26.11 by @benceruleanlu in https://github.com/comfyanonymous/ComfyUI/pull/9809
* Add noise augmentation to hunyuan image refiner. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9831
* Fix hunyuan refiner blownout colors at noise aug less than 0.25 by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9832
* Set default hunyuan refiner shift to 4.0 by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9833
* API nodes: add kling-v2-1 model to the KlingStartEndFrame node by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9630
* [V3] convert Minimax API nodes to the V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9693
* [V3] convert WanCameraEmbedding node to V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9714
* [V3] convert Cosmos nodes to V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9721
* [V3] convert nodes_cond.py to V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9719
* [V3] convert CFG nodes to V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9717
* [V3] convert Canny node to V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9743
* [V3] convert Moonvalley API nodes to the V3 schema by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9698
* Better way of doing the generator for the hunyuan image noise aug. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9834
* Enable Runtime Selection of Attention Functions by @Kosinkadink in https://github.com/comfyanonymous/ComfyUI/pull/9639
* Hunyuan refiner vae now works with tiled. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9836
* Support wav2vec base models by @kijai in https://github.com/comfyanonymous/ComfyUI/pull/9637
* Cleanup. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9838
* HunyuanImage: Remove single quote pattern to avoid wrong text matches by @KimbingNg in https://github.com/comfyanonymous/ComfyUI/pull/9842
* Add support for Chroma Radiance by @blepping in https://github.com/comfyanonymous/ComfyUI/pull/9682
* Changes to the previous radiance commit. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9851
* Make ModuleNotFoundError ImportError instead by @Kosinkadink in https://github.com/comfyanonymous/ComfyUI/pull/9850
* Add that hunyuan image is supported to readme. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9857
* Support the omnigen2 umo lora. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9886
* Fix python -O and BatchedBrownianTree w0 handling by @blepping in https://github.com/comfyanonymous/ComfyUI/pull/9884
* Add encoder part of whisper large v3 as an audio encoder model. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9894
* Reduce Peak WAN inference VRAM usage by @rattus128 in https://github.com/comfyanonymous/ComfyUI/pull/9898
* Support the 1.7B HuMo model. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9903
* Support the HuMo 17B model. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9912
* Enable fp8 ops by default on gfx1200 by @0xDELUXA in https://github.com/comfyanonymous/ComfyUI/pull/9926
* Fix image sharpen node by @jetjodh in https://github.com/comfyanonymous/ComfyUI/pull/9932
* Do padding of audio embed in model for humo for more flexibility. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9935
* Bump frontend to 1.26.13 by @christian-byrne in https://github.com/comfyanonymous/ComfyUI/pull/9933
* Basic WIP support for the wan animate model. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9939
* fix(api_nodes): reduce default timeout from 7 days to 2 hours by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9918
* fix(seedream4): add flag to ignore error on partial success by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9952
* Update WanAnimateToVideo to more easily extend videos. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9959
* Add inputs for character replacement to the WanAnimateToVideo node. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9960
* [Reviving #5709] Add strength input to Differential Diffusion by @Kosinkadink in https://github.com/comfyanonymous/ComfyUI/pull/9957
* Fix LoRA Trainer bugs with FP8 models. by @KohakuBlueleaf in https://github.com/comfyanonymous/ComfyUI/pull/9854
* Lower wan memory estimation value a bit. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9964
* Set some wan nodes as no longer experimental. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9976
* Support for qwen edit plus model. Use the new TextEncodeQwenImageEditPlus node. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9986
* Add offset param to `/history` endpoint by @christian-byrne in https://github.com/comfyanonymous/ComfyUI/pull/9977
* Fix bug with WanAnimateToVideo node. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9988
* Fix bug with WanAnimateToVideo. by @comfyanonymous in https://github.com/comfyanonymous/ComfyUI/pull/9990
* update template to 0.1.86 by @comfyui-wiki in https://github.com/comfyanonymous/ComfyUI/pull/9998
* feat(api-nodes): add wan t2i, t2v, i2v nodes by @bigcat88 in https://github.com/comfyanonymous/ComfyUI/pull/9996

## New Contributors
* @KimbingNg made their first contribution in https://github.com/comfyanonymous/ComfyUI/pull/9842
* @rattus128 made their first contribution in https://github.com/comfyanonymous/ComfyUI/pull/9898
* @0xDELUXA made their first contribution in https://github.com/comfyanonymous/ComfyUI/pull/9926
* @jetjodh made their first contribution in https://github.com/comfyanonymous/ComfyUI/pull/9932

**Full Changelog**: https://github.com/comfyanonymous/ComfyUI/compare/v0.3.59...v0.3.60