Description
Watching a foreign-language video with no subtitles is the worst: a line of dialogue you cannot follow goes by, pausing to look it up breaks the immersion, and pushing on loses the plot. Raw live streams, game dialogue with no localisation yet, foreign films — the wall is always the language.
Mimi (“ear” in Japanese) is built for that. It captures the audio your system is playing, turns it into live subtitles and translates them, overlaid on screen — and because it listens to the system output, anything playing gets subtitles regardless of the player, browser or game. Translation between Simplified Chinese, English and Japanese, depending on the provider you use.
Subtitles show as an overlay you can move, resize, collapse and pause, with click-through so it does not block your actions, and an Immersive Mode a shortcut away. On privacy it stays clean: no account, no microphone, no screen capture, and no saved audio or subtitles — audio goes only to the provider you configure.
It uses your own provider API credentials, stored in the OS credential store. Written in Rust with Tauri, MIT licensed, for Apple silicon macOS 13+ and Windows x64.
Live subtitles and translation: Captures system output audio, transcribes it live and translates it; source and target languages and quality modes depend on the provider.
Captures system audio: It listens to the system output rather than one app, so player, browser, live-stream and game audio can all be subtitled.
Switchable service configs: Save credentials for several translation services and switch between them without re-entering them.
Flexible overlay: Move, resize, collapse and pause the overlay, enable click-through so it does not block input, and toggle Immersive Mode with a shortcut.
Clean on privacy: No account, no microphone, no screen capture, and no saved audio or subtitles; audio goes only to the provider you set.
In-app updates: Check and install updates from Settings rather than in the background.
Mimi (“ear” in Japanese) is built for that. It captures the audio your system is playing, turns it into live subtitles and translates them, overlaid on screen — and because it listens to the system output, anything playing gets subtitles regardless of the player, browser or game. Translation between Simplified Chinese, English and Japanese, depending on the provider you use.
Subtitles show as an overlay you can move, resize, collapse and pause, with click-through so it does not block your actions, and an Immersive Mode a shortcut away. On privacy it stays clean: no account, no microphone, no screen capture, and no saved audio or subtitles — audio goes only to the provider you configure.
It uses your own provider API credentials, stored in the OS credential store. Written in Rust with Tauri, MIT licensed, for Apple silicon macOS 13+ and Windows x64.
Features
Live subtitles and translation: Captures system output audio, transcribes it live and translates it; source and target languages and quality modes depend on the provider.
Captures system audio: It listens to the system output rather than one app, so player, browser, live-stream and game audio can all be subtitled.
Switchable service configs: Save credentials for several translation services and switch between them without re-entering them.
Flexible overlay: Move, resize, collapse and pause the overlay, enable click-through so it does not block input, and toggle Immersive Mode with a shortcut.
Clean on privacy: No account, no microphone, no screen capture, and no saved audio or subtitles; audio goes only to the provider you set.
In-app updates: Check and install updates from Settings rather than in the background.
