How to serve your ComfyUI model behind an API endpoint

In this guide, we’ll deploy image generation pipelines built with ComfyUI behind an API endpoint so they can be shared and used in applications. To serve the model pipeline in production, we’ll export the ComfyUI project in an API format, then use Truss for packaging and deployment.

[chaoslawine]

Thanks for the video, exactly what I'm looking for

[chrisgreenwell]

this is very cool. What I would be interested in is the ability to run this locally for a app like Stable Projectorz to connect to, it would need to be able to query the existing models and controlnets for selection from installed ones. Is there a way to do this with this type of system ?

[vrflowrrr]

Getting this:
You have to log in to perform the request
<Server response: {'errors': [{'message': 'You have to log in to perform the request', 'locations': [{'line': 3, 'column': 13}], 'path': ['models'], 'extensions': {'code': 'UNAUTHENTICATED_ACCESS'}}], 'data': {'models': None}}>

[Takio]

too sad you guys are double as expensive than others which are already overpriced :/

[vedanthbora]

Im facing an issue deploying a model form one of truss exmaples. Im trying to deploy the comfy-ui example.

The base image in the config.yaml file points to



this image basically has an outdated version of the comfy ui due to which things break in prod.

Ive noticed many folks face this problem.

Can you please help me out here