ComfyUI 0.29 adds streaming video, GPT 5.6, Claude Opus 5, and new partner nodes
pytraveler · reddit · 2026-07-29
ComfyUI 0.29 is a substantial release focused on three areas: faster video handling, broader model support, and performance work.
- Video: streaming transcoding now processes clips on the fly instead of buffering entire files into RAM, which lowers memory use for long videos. Video support also landed in the trainer and image-processing nodes.
- Models / partner nodes: new or updated integrations include GPT 5.6, Gemini 3.5 Flash, Gemini-Omni fixes, Claude Opus 5, ByteDance audio-1.0-multilingual, HeyGen avatar and video-translate nodes, Recraft V4.1, and deprecation of Runway Gen3a.
- Local models: support expanded for JoyImageEdit, Gemma4 12B, MageFlow, Uni3C ControlNet, Anima, FreSca, Krea 2, and LTX improvements.
- Performance: optimized RMS RoPE paths, more INT8 optimizations, and a faster latent upsampler via ModelPatcherDynamic.
- System/API: the background memory default rises to 128 GB, AGENTS.md is added to CODEOWNERS, and partner integrations now get Job ID / credit headers.
More from coding & agent
- Replication finds agent experience distillation preserves 44.1% of ICL gains on SWE tasks — burny_tech · 2026-07-29
- Grok Build adds MCP CLI controls, instant cold start and a /delete session command — mark_k · 2026-07-29
- LLM coding workflow adds a review gate to stop every retrieved method from being implemented — hypergraphr · 2026-07-29
- Yangqing Jia and Andrew Ng Announce AI Startups: Engineering Agents and Personalized Education — 机器之心 · 2026-07-29
- A GPT code-fix screenshot turns into a classic overconfident AI meme — hrishioa · 2026-07-29
- The sane middle ground for AI coding is to build your own workflow and iterate weekly — paulabartabajo_ · 2026-07-29