ComfyUI
Node-based stable-diffusion UI for building custom image-generation workflows visually.
ComfyUI keeps day-0 model integrations coming and now opens an agent control surface
◆Recent moves
- 4d ago
Comfy MCP: Turn your agent into a creative technologist
⚡ SPARKComfy MCP exposes the entire ComfyUI ecosystem to agents like Claude, Codex, Hermes, and Cursor. It marks a shift from a human-operated node editor toward a backend that AI assistants can drive directly, layered on top of the existing model-integration cadence.
View source ↗ - 8d ago
Seedance 2.0 Mini and 4K is now available in ComfyUI
Seedance 2.0 Mini lands and full Seedance 2.0 gains 4K output. A typical day-0 model integration that broadens the video-generation options available inside workflows.
View source ↗ - 8d ago
HappyHorse 1.1 is now available in ComfyUI
HappyHorse 1.1 adds audio-native video generation with dialogue, sound effects, and multi-character consistency inside workflows. Another model integration extending ComfyUI's video coverage into audio-aware generation.
View source ↗ - 9d ago
Krea 2 Open-Source Models are now available in ComfyUI
Krea 2 open-source checkpoints (RAW and Turbo) become available in ComfyUI. Fits the steady pattern of integrating open-weights models as they ship.
View source ↗ - 10d ago
The tool that expands my art: Xindi Zhang's Oscar-shortlisted thesis, built in ComfyUI
A creative-showcase post about a USC thesis built in ComfyUI. Community storytelling, not a product change.
View source ↗ - 11d ago
I built a native Comfy Cloud mobile app on nothing but the public API
A community post about one developer building a Comfy Cloud mobile app on the public API. Illustrative of API reach but not itself a ComfyUI release.
View source ↗