hero8152/infinite-canvas — explained in plain English
Analysis updated 2026-05-18
Generate multiple AI images or videos in one batch using a looped prompt setup
Connect ComfyUI to external API providers and ModelScope models without leaving the canvas
Build custom ComfyUI workflows with adjustable inputs and parameters
| hero8152/infinite-canvas | cslawyer1985/claude-for-legal-zh | opendrivelab/simscale | |
|---|---|---|---|
| Stars | 266 | 264 | 263 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | moderate | hard |
| Complexity | 3/5 | 3/5 | 5/5 |
| Audience | vibe coder | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires an existing ComfyUI install plus an API key for an external model provider.
Infinite Canvas is a ComfyUI extension that connects an infinite visual workspace to outside AI services. It works with ComfyUI itself, with API calls to hosted AI providers, and with ModelScope, a model hosting platform, so you can pull in different AI models and use them inside your canvas without leaving ComfyUI. The project has been updated frequently. Recent changes fixed installation errors, added an easier way to set up API connections directly from a webpage instead of editing config files, and made it possible to add models with one click. The LLM node now accepts image input for reverse engineering tasks, and you can test it against vision language models hosted on ModelScope. There is also a language switch between Chinese and English, support for custom ComfyUI workflows with your own inputs and adjustable parameters, and a video generation feature. A loop component lets a single node run multiple times in a row, useful for batch tasks. For example, you could use another AI model to write ten product selling point descriptions, then use the loop node to feed each one into an API call that generates a matching image, producing ten images in one pass instead of doing it one at a time. The project also added a button to verify which protocol your chosen API platform uses, since it now supports both the OpenAI style protocol and an asynchronous protocol where failed generations are not charged. The README points to a video tutorial for a full walkthrough of setup and use. It also mentions a specific API provider the author recommends after other API sites shut down, including a discount code for new signups. For simpler installation, a trimmed down Python bundle is available: you run one batch file to install dependencies, then another to start the tool. This project would suit someone already using ComfyUI who wants to add flexible calls to external AI APIs and ModelScope models directly inside their existing image or video generation workflows, rather than switching between separate tools.
A ComfyUI extension that lets you call external AI APIs and ModelScope models directly inside an infinite visual workspace, with loop tools for batch image and video generation.
Mainly Python. The stack also includes Python, ComfyUI.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
Verify against the repo before relying on details.