mirumoji

July 14, 2026 ยท View on GitHub

PyPI License Docs Docker Pulls Release Action GitHub Downloads (all assets, all releases)

An open-source, self-hosted Japanese Immersion Toolkit

Drop in a video, an anime episode, a drama, or an audio clip and Mirumoji gives you clickable tokenized subtitles with instant dictionary lookups whisper-powered subtitle generation, clip saving, and Anki export. All running locally in Docker, with optional cloud GPU and LLM features

Mirumoji Demo

Quickstart

Mirumoji runs as a local Docker Compose Application

The quickest way to start is to download the Desktop Launcher for your platform and follow the Setup Walkthrough

Alternatives

You can also setup Mirumoji Manually, or through its CLI

Features

Interactive Player Load Videos + Subtitles For Clickable, Tokenized Japanese Lines With Dictionary Pop-Ups
Transcription Generate Subtitles + Transcribe Audio with faster-whisper, on a Local / Cloud GPU
Dictionary Wildcard dictionary search (kotobase) + a paste-in text analyzer with furigana
LLM Optionally Use Gemini / Claude / GPT API Keys (Or Your Own Local LLM Server) To Generate Sentence Breakdowns With Fully Customizable Prompts
Clips + Anki Save Video Segments With Their Word Breakdowns And Export Them As An Anki Deck
Profiles Keep Generated Files, Transcripts, Clips, and LLM Templates Organized Per Profile On The Server
Launcher Easily Run The Docker Compose Application Via The CLI Or The Desktop Launcher

Help

Documentation Guides + Full Backend / Frontend API Reference
Setup Guide CLI + GUI + Manual setup
Using a GPU Local GPU x Modal Cloud GPU
Changelog What Changed Between Versions
Demo Test A Limited Preview

Contributing

  • All contributions are welcome

  • See CONTRIBUTING for more information on how to contribute