The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Commits on master
[Partner Nodes] feat(HeyGen): add HeyGen Voice 1.0 instant clone and text-to-speech nodes (#17043)
a2440bc Alexander Piskun committed 1 month ago
[Partner Nodes] Renew partner-node sign-in tokens in API nodes (#16943)
8f04fd6 Dante committed 1 month ago
CPU device doesn't support fp8, nvfp4, mxfp8 or int8. (#17170)
d4056d5 comfyanonymous committed 1 month ago
Fix text generation not properly cleaning itself up. (#17169)
b352238 comfyanonymous committed 1 month ago
Make NAF neighborhood attention work per LR cell (#17155)
446ef4e Jukka Seppänen committed 1 month ago
add support for loading embeddings models onto Qwen3.5 (#17159)
872eb6a Silver committed 1 month ago
Fix Qwen think-block templates in TextGenerate (#17121)
ce1e922 Jukka Seppänen committed 1 month ago
[Partner Nodes] chore(Quiver): remove retired Arrow 1.x models (#17102)
fc04da9 Alexander Piskun committed 1 month ago
Better documentation about the comfy quant format. (#17078)
5a66589 comfyanonymous committed 1 month ago
[Partner Nodes] feat(Krea): add Krea 3 Image node (#17067)
a5e95e6 Alexander Piskun committed 1 month ago