Skip to content

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

Disable workaround on newer cudnn. (#10807)
comfyanonymous committed 1 year ago
cb96d4d
feat(api-nodes): add Topaz API nodes (#10755)
Alexander Piskun committed 1 year ago
394348f
Fix workflow name. (#10806)
comfyanonymous committed 1 year ago
7601e89
convert hunyuan3d.py to V3 schema (#10664)
Alexander Piskun committed 1 year ago
6a1d3a1
ComfyUI 0.3.70
comfyanonymous committed 1 year ago
b5c8be8
feat(api-nodes): add new Gemini model (#10789)
Alexander Piskun committed 1 year ago
24fdb92
Fix hunyuan 3d 2.0 (#10792)
comfyanonymous committed 1 year ago
d526974
47bfd5a
fdf49a2
Update README with new portable download link (#10778)
comfyanonymous committed 1 year ago
f41e5f3
Add release workflow for NVIDIA cu126 (#10777)
comfyanonymous committed 1 year ago
27cbac8
ComfyUI version 0.3.69
comfyanonymous committed 1 year ago
3d0003c
Change ROCm nightly install command to 7.1 (#10764)
comfyanonymous committed 1 year ago
7d61033
Add left padding support to tokenizers. (#10753)
comfyanonymous committed 1 year ago
bd01d9f
Fix custom nodes import error. (#10747)
comfyanonymous committed 1 year ago
443056c
Better instructions for the portable. (#10743)
comfyanonymous committed 1 year ago
1ef328c
flux: reduce VRAM usage (#10737)
rattus committed 1 year ago
94c298f
add PR template for API-Nodes (#10736)
Alexander Piskun committed 1 year ago
f91078b
Quantized Ops fixes (#10715)
contentis committed 1 year ago
3b3ef9a
8b0b93d
qwen: reduce VRAM usage (#10725)
rattus committed 1 year ago
1c7eaec