comfyui support
#2
by ZKong - opened
thanks
ComfyUI support is on our radar. In the meantime, you can create a custom node that calls the Scenema Audio REST API locally. The Docker service runs on localhost:8000 and the node just builds the
XML prompt from your inputs and POSTs to /generate.
See the GitHub README for setting up the Docker service. Once it's running, any ComfyUI node that makes an HTTP call to localhost:8000/generate will work.
We may publish an official ComfyUI node package if there's enough interest.
Great. what i mean is native comfyui workflow, which cpu offload is very good. LTX can run in less than 6G vram in native workflow
Oh not yet. That's definitely something we're thinking about though.