Verified project record

linyqh/NarratoAI

NarratoAI is an open-source, LLM-driven pipeline for creating narrated film and TV commentary videos, combining AI script writing, automated video editing, dubbing, and subtitles. It can be self-hosted locally or used through an optional hosted cloud service.

Project overview

It consolidates the entire commentary-video workflow—script writing, editing, dubbing, and subtitles—into one automated pipeline, and supports local open-source deployment with voice cloning across multiple TTS engines including a local option for Apple Silicon.

Project type: Video
Use cases: Documents & Office · Video Creation
Deployment: Refer to project documentation
License: MIT

Best for

Creators and general users who make short narrated film/TV commentary videos and want AI-generated copy with automated editing, dubbing, and subtitles.

Key capabilities

Uses an LLM to generate narration copy for film/TV commentary videos.
A one-stop automated pipeline that edits video and generates dubbing and subtitles.
Supports voice cloning and multiple TTS engines, including a local MLX voice-clone engine for Apple Silicon.

Limitations and risks

Script generation requires an LLM accessed via an OpenAI-compatible API; script writing depends on this external model service.
External service usage is optional but present: script generation calls an external LLM API, and an optional hosted cloud version exists at NarratoAI.cn. Local deployment means the pipeline is self-hosted, not that all processing stays local.

Getting started

Clone the repository with 'git clone https://github.com/linyqh/NarratoAI.git', run 'docker compose up -d', then open http://localhost:8501 in a browser to use the web UI. A 4-core CPU, 8GB RAM, Windows 10/11 or macOS 11.0+, and Python 3.12+ are recommended; a GPU is not required.

Evidence and sources

GitHub project description
README
README
README
README: git clone https://github.com/linyqh/NarratoAI.git

Source updated at: 2026-10-25T18:08:18Z

Open original GitHub project