hero8152/Infinite-Canvas

What it solves

Infinite-Canvas provides a unified, infinite canvas interface for interacting with various AI models and workflows. It solves the problem of fragmented AI tools by integrating multiple API protocols and local installations of ComfyUI, allowing users to manage generative AI tasks in a visual, collaborative environment.

How it works

The project acts as a a visual frontend that supports a wide range of integrations:

• API Support: It integrates with OpenAI-compatible APIs, Gemini, and Modelscope, as well as specific platforms like RunningHub and Jimeng (即梦) CLI for text-to-image, image-to-image, and video generation.
• Local Integration: It can launch and connect to local ComfyUI instances on a local area network (LAN).
• Visual Tools: It includes features for image expansion, 360-degree panoramic previews, and video frame extraction.
• Ecosystem: It provides a Chrome extension for collecting materials and a Photoshop plugin for direct canvas integration.

Who it’s for

Digital artists, AI prompt engineers, and developers who need a visual workspace to orchestrate complex AI generative workflows across different model providers.

Highlights

• Multi-Protocol Support: Compatible with almost all OpenAI-style APIs, Gemini, and Modelscope.
• ComfyUI Integration: Direct support for launching and mapping workflows from ComfyUI.
• Collaborative Features: Supports multi-person collaboration, shared canvases, and one-click layering.
• Extended Tooling: Includes a Chrome material collection plugin and a PS direct-connect plugin.