Mimi is a real-time subtitle and translation application that converts computer or microphone audio into readable captions. It helps users understand foreign-language videos, livestreams, online classes, and spoken conversations without relying on embedded subtitles. The application supports bilingual captions or separate original-language and translated text displays. A floating subtitle overlay can be repositioned and customized with adjustable fonts, colors, and immersive viewing options. Users can capture audio from specific applications on supported desktop systems or combine system and microphone inputs. Speech recognition and translation integrate with providers such as OpenAI, Gemini, and xAI through user-supplied API credentials. Optional recording and transcript-saving tools export text as TXT or Markdown and audio as WAV. Mimi supports Windows, macOS, Linux, and Android.
Features
- Real-time speech transcription and translation for videos, livestreams, meetings, and online classes
- Customizable floating subtitles with bilingual, original-language, or translated-text display modes
- Application-specific audio capture on supported systems with microphone and system audio input options
- Integration with OpenAI, Gemini, xAI, and compatible speech recognition or translation services
- Optional transcript recording with TXT and Markdown exports alongside WAV audio capture
- Cross-platform compatibility with immersive subtitle overlays and click-through interaction controls