Toonflow-app is an open-source desktop-oriented AI tool that turns stories into animated short dramas with AI scripting, storyboarding, and video generation. It provides a multi-agent workflow, local memory, and TypeScript-based plugin system, with Docker and cloud deployment options.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
Toonflow-app is an open-source one-stop AI short-drama tool that converts stories and scripts into animated short dramas. It includes AI scripting, storyboarding, character and video generation, and a cross-platform desktop app for local deployment.
How it works
Key features include an infinite-canvas style workspace organizing scripts, characters, storyboards, assets, and video nodes; a three-layer Agent system for decision, execution, and supervision; a local ONNX-based memory system for cross-session recall; a programmable supplier system enabling TypeScript logic in a settings center without source changes or restarts; chapter-events graph to drive adaptation; and Skill files that externalize core prompts for ScriptAgent and ProductionAgent.
Getting started
Quick start
- Install and log in (default account: admin / admin123).
- Configure model vendors in Settings (text/image/video models).
- Create a project and import the source, run chapter-event extraction.
- Use ScriptAgent to generate story skeletons and structure.
- Switch to ProductionAgent to organize storyboards, assets, and video nodes.
- Refine storyboards as nodes and export the video.
Local install and Docker
- Local install via releases on Windows/Linux/macOS. First login details are provided in the README.
- Docker/deploy options include local docker-compose build and a manual docker run example, with environment variable notes and port mappings (default port 10588).
- Environment variables: NODE_ENV, PORT, OSSURL.
Cloud deployment
- Steps for cloning from GitHub or Gitee, installing dependencies, and building, followed by PM2 process management.
- PM2 configuration example sets NODE_ENV to prod, PORT to 10588, and OSSURL to http://127.0.0.1:10588/.
Recent releases
- v1.1.8 ToonFlow (2026-06-08): Windows x64 setup and Windows ARM64 setup files provided.
- v1.1.7 ToonFlow (2026-05-01): Windows x64 setup and Windows ARM64 setup files provided.
- v1.1.6 ToonFlow (2026-04-19): Windows x64 setup and Windows ARM64 setup files provided.
- v1.1.5 ToonFlow (2026-04-14): Windows x64 setup and Windows ARM64 setup files provided.
- v1.1.4 ToonFlow (2026-04-11): Windows generic and x64 setup files provided.
Traction
- Stars: 13382; Forks: 2427; Open issues: 3. Language: TypeScript. License: Apache-2.0.
Behind the repo
- Maintains multi-language READMEs and localization for simplified Chinese, traditional Chinese, English, Thai, Vietnamese, Japanese, and Russian. Tech stack includes Node.js, TypeScript, Express, SQLite, Vercel AI SDK, ONNX, Socket.IO, Electron, and Docker.
Caveats
- Prerequisites include AI service interfaces for language, image, and video models; specific external services are referenced (e.g., token service and cloud providers).
- The repository includes Docker deployment steps, cloud deployment guidance, and environment variable configurations; production deployment requires setting up models, tokens, and PM2 process management.
- The README notes about macOS security settings and localization considerations; no explicit licensing restrictions beyond Apache-2.0 are listed in the quoted content.