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
Specify in readme that we only support pytorch 2.4 and up. (#11512)
d9a76cf comfyanonymous committed 10 months ago
api-nodes: use new custom endpoint for Nano Banana (#11311)
f4f44bb Alexander Piskun committed 11 months ago
Make denoised output on custom sampler nodes work with nested tensors. (#11471)
33aa808 comfyanonymous committed 11 months ago
extend possible duration range for Kling O1 StartEndFrame node (#11451)
c176b21 Alexander Piskun committed 11 months ago
Add node to create empty latents for qwen image layered model. (#11460)
91bf6b6 comfyanonymous committed 11 months ago
chore(api-nodes): by default set Watermark generation to False (#11437)
0899012 Alexander Piskun committed 11 months ago
Only apply gemma quant config to gemma model for newbie. (#11436)
fb478f6 comfyanonymous committed 11 months ago
ZImageFunControlNet: Fix mask concatenation in --gpu-only (#11421)
767ee30 rattus committed 11 months ago
Allow enabling use of MIOpen by setting COMFYUI_ENABLE_MIOPEN=1 as an env var (#11366)
cc4ddba BradPepersAMD committed 11 months ago
Support loading Wan/Qwen VAEs with different in/out channels. (#11405)
e4fb3a3 comfyanonymous committed 11 months ago