comfyui
https://github.com/comfyanonymous/comfyui
Python
The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported24 Subscribers
Add a CodeTriage badge to comfyui
Help out
- Issues
- Fix Generate LTX2 Prompt leaking reasoning chain for Qwen3.5
- docs: fix dead in-page anchor in README Apple Silicon install section
- Add NVIDIA Easypanel deployment configuration
- Add optional ROCmFPX Qwen text encoders
- INT8 ConvRot fails on Turing GPUs for M=97..127 despite SM 7.5 support
- Improve default inpainting workflow by compositing result over original image with ImageCompositeMasked
- Qwen3-VL Generate Text aborts with "scatter gather kernel index out of bounds" on the second generate() in one node execution
- [Qwen Image 2.1] Hard crash in prefetch/staging-buffer path on Windows multi-GPU (Fatal Python error: Aborted)
- Qwen-Image 2.1: default sampling uses a fixed mu=0.69 for every resolution, so the out-of-the-box schedule no longer matches the model's published dynamic-shift scheduler
- Qwen-Image-2.1 VAE encode is broken on MPS: encode→decode round-trip returns 6.6 dB PSNR (49.1 dB on CPU), silently corrupting every image-edit workflow
- Docs
- Python not yet supported