NarratoAI
Open-source tool that uses LLMs to write narration scripts and auto-edit video clips.
- Stars
- 11.7k
- Forks
- 1.6k
- Licence
- MIT
- Language
- Python
- Last update
- 13 days ago
- Type
- App
What it is
NarratoAI is a tool for making commentary-style videos. It uses large language models to write narration scripts, then handles clip editing, voiceover, and subtitle generation in one workflow. The README describes both a local open-source version and a hosted cloud version.
- Generate a narration script for a video clip using an LLM
- Automatically cut and assemble video footage with narration
- Generate voiceover and subtitles, including transcription with Fun-ASR per the release notes
- Export drafts to CapCut (剪映) per the roadmap
Great for
- Creators who make commentary or short-drama mashup videos
- People comfortable installing Python tools or using Docker
- Users who want to try AI voiceover, including voice cloning engines the README lists
- Users who want an offline-leaning local setup with their own API keys
Not ideal for
- People who want a simple mobile or no-setup experience, since the README centers on Windows, macOS, Docker, or a Python setup
- Users who need a verified, fully offline workflow, since the README requires configuring API keys for LLM and TTS providers
- Anyone who needs guaranteed stability, since the README lists frequent releases and a large 0.8.x version jump without detailed change notes
How to get it
- 1
Run with Docker
You'll need: Docker · Python 3.12+ · uv
Clone the repository
git clone https://github.com/linyqh/NarratoAI.gitEnter the project folder
cd NarratoAIStart with Docker Compose
docker compose up -d - 2
Run it
- Run the Docker command or the local uv setup from the README
- Open http://localhost:8501 in a browser
- Add your API key settings in config.toml (local setup) or the app's settings
- Upload footage and run the auto-edit workflow
- 3
Get the most out of it
- Use the official documentation linked in the README for detailed setup
- Review the 0.8.x release notes for new features such as Sonilo AI music and IndexTTS engines
- Try the TwelveLabs video understanding option by setting vision_llm_provider in config
- Use the short-drama mashup mode for combining multiple clips, per the changelog
Downloads
Strengths
- Covers script writing, editing, voice, and subtitles in one pipeline
- MIT license with a clear open-source version and active release history
- Multiple install paths: prebuilt packages, Docker, or local uv setup
Trade-offs
- Setup requires API keys from model providers, and some features depend on specific providers
- The README is largely in Chinese and points to Chinese-language docs and community links
- Some engines are macOS-only or lack Windows versions, as noted in the v0.8.6 table
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| NarratoAI (this) | 11.7k | App | intermediate | 13 days ago | Creators who make commentary or short-drama mashup videos |
| MoneyPrinterTurbo | 137k | App | intermediate | 4 days ago | Creators who want quick draft short videos from a topic |
| MoneyPrinter | 14.1k | App | intermediate | 7 months ago | Hobbyists who want to experiment with automated short-form video creation |
| ida-pro-mcp | 13.3k | App | intermediate | 1 months ago | Reverse engineers who already use IDA Pro 8.3 or newer and want an AI assistant to help with analysis |
| PocketFlow | 11.3k | Library | intermediate | 3 months ago | Developers who want a small, readable LLM framework they can inspect in full |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.