Repolode
Submit
WEIFENG2333

WEIFENG2333/VideoCaptioner

An LLM-powered desktop and CLI tool for video subtitles: speech recognition, sentence segmentation, correction, translation and burning subtitles into video.

Open-source alternative to:Subtitle EditAegisub

View on GitHubWebsite
Stars
16,297
Forks
1,415
Open issues
214
Last commit
last month

What is it?

VideoCaptioner handles the full subtitle workflow for video files: it transcribes speech, splits the text into readable subtitle lines, optimizes wording with an LLM, translates the result, and synthesizes subtitles back into the video. It ships as both a command-line tool and a desktop GUI, installable with pip.

Several steps work without any configuration. Transcription can use the free Bijian and Jianying engines, and translation can use Bing or Google. LLM-based optimization and LLM translation require an API key for any OpenAI-compatible service, with settings configurable via the GUI, CLI or environment variables.

The project also provides a Claude Code skill, so AI coding assistants can call it directly to process videos. It can download videos from platforms such as YouTube and Bilibili as a starting point for processing.

Key features

  • 01Speech transcription with multiple engines, including free Bijian and Jianying options
  • 02LLM-based subtitle segmentation and proofreading
  • 03Subtitle translation via LLM, Bing or Google
  • 04Burning subtitles into video as soft or hard subtitles
  • 05Full pipeline command that chains transcription, optimization, translation and synthesis
  • 06Desktop GUI and CLI with layered configuration

Use cases

  • Generating subtitles for videos without writing them by hand
  • Translating a video's subtitles into another language
  • Batch processing subtitle files from the command line
  • Letting an AI coding assistant handle video subtitle tasks through a Claude Code skill

Quick start

  1. 1

    Install the CLI and GUI from PyPI

    pip install videocaptioner
  2. 2

    Open the desktop GUI

    videocaptioner-gui
  3. 3

    Or transcribe a video from the command line (free, no API key)

    videocaptioner transcribe video.mp4 --asr bijian

Things to know

  • LLM-based features such as subtitle optimization and LLM translation require your own API key from an OpenAI-compatible provider.
  • Licensed under GPL-3.0, so derivative distributions must also be open source under GPL terms.

Topics

#subtitle#video#speech-recognition#llm#translation#whisper#cli#ai#translate#video-subtile

FAQ

What is VideoCaptioner?
An LLM-powered desktop and CLI tool for video subtitles: speech recognition, sentence segmentation, correction, translation and burning subtitles into video. VideoCaptioner handles the full subtitle workflow for video files: it transcribes speech, splits the text into readable subtitle lines, optimizes wording with an LLM, translates the result, and synthesizes subtitles back into the video. It ships as both a command-line tool and a desktop GUI, installable with pip.
Is VideoCaptioner free to use? What license does it have?
VideoCaptioner is open source under the GPL-3.0 license. Check the license terms before using it in commercial products.
Is VideoCaptioner still maintained?
Updated within 6 months. The last commit to VideoCaptioner was on Sep 12, 2026.
What can I use VideoCaptioner for?
Generating subtitles for videos without writing them by hand Translating a video's subtitles into another language Batch processing subtitle files from the command line Letting an AI coding assistant handle video subtitle tasks through a Claude Code skill
How do I install VideoCaptioner?
Install the CLI and GUI from PyPI: pip install videocaptioner Open the desktop GUI: videocaptioner-gui Or transcribe a video from the command line (free, no API key): videocaptioner transcribe video.mp4 --asr bijian
Where can I get VideoCaptioner?
The source code and installation instructions are on GitHub: https://github.com/WEIFENG2333/VideoCaptioner

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects