An open-source infinite canvas workspace for AI creation, integrating AI image generation, reference image editing, video generation, an Agent assistant, canvas orchestration, conversational creation, prompt libraries, and asset management. Supports visual creation workflows and multi-Agent collaboration. Compatible with the OpenAI API ecosystem, with support for channels such as chatgpt2api, grok2api, flow2api, and newapi.
infinite-canvas is an open-source visual workspace for AI image creation, combining an infinite canvas with image generation, reference-image editing, video generation, chat-based assistance, prompt storage, asset management, and multi-agent workflows. It is for exploring and iterating on visual concepts through a graphical creative process, with access to OpenAI-compatible model services. The catalogue contains skills, an MCP integration, and an instruction for working with the project.
This repository also configures its own agents. See what infinite-canvas tells them →
MCP server Claude CodeCodexCursor +2
A local bridge between a web-based drawing canvas and coding agents such as Codex or Claude Code.