AI Canvas (huobao-canvas) provides a visual, node-based interface for orchestrating AI-powered creative workflows, primarily focused on media generation. It addresses the complexity of chaining AI models for tasks like text-to-image and image-to-video by offering an intuitive, Vue Flow-based infinite canvas. This tool is designed for engineers, researchers, and power users seeking a streamlined way to build and manage AI content generation pipelines, offering benefits like automated workflow execution and AI-assisted prompt refinement.

How It Works

The project leverages Vue Flow to create a highly interactive, infinite canvas where users can visually construct AI workflows. Core components include distinct nodes for defining prompts, configuring image and video generation parameters, and displaying generated media. An AI intent analysis engine interprets user input to automatically select and orchestrate the appropriate workflow (e.g., text-to-image, text-to-image-to-video, storyboard generation), creating nodes and executing them sequentially. This approach simplifies complex AI task management by abstracting the underlying API calls and model interactions into a visual, manageable system.

Quick Start & Requirements

• Environment: Node.js >= 18.
• Installation: Clone the repository, install dependencies using pnpm install (or npm install), and start the development server with pnpm dev (or npm run dev).
• Configuration: Requires setting up API Base URL and API Key for OpenAI-compatible APIs via a settings icon in the UI.
• Links: Project repository: https://github.com/chatfire-AI/huobao-canvas

Highlighted Details

• Visual node-based orchestration powered by Vue Flow.
• Supports text-to-image and image-to-video generation workflows.
• Includes an AI feature for optimizing and refining prompts.
• Automatic workflow execution based on AI intent analysis of user input.
• Local project storage with multi-project management and undo/redo functionality.
• Offers both dark and light theme options.

Maintenance & Community

The project follows standard GitHub contribution guidelines for Issues and Pull Requests. Community interaction appears to be facilitated via WeChat, as indicated by a QR code in the README, though direct links to Discord or Slack are not present.

Licensing & Compatibility

The project is released under the MIT License. This permissive license generally allows for commercial use, modification, and distribution, including integration within closed-source applications.

Limitations & Caveats

The core AI functionality relies on external OpenAI-compatible API endpoints, meaning users must provide their own API keys and base URLs. The effectiveness of the automatic workflow execution and AI intent analysis may depend on the specific AI models used and the clarity of user prompts.