Memo AI
An offline transcription app for Windows and macOS that turns recordings into text and translated subtitles, with local speaker labels and GPU acceleration.
Memo AI turns audio and video into transcripts, translated subtitles and notes for people reviewing interviews, meetings or podcasts. Speech recognition works offline, and speaker identification runs locally.
- Transcription and translation: Process local recordings, YouTube videos and podcasts, with language support including Chinese, English and Japanese and context-aware translation for long subtitles.
- Speaker labels: Distinguish participants in conversations so you can follow who said what.
- Playback review: Read subtitles during playback, display key points in floating notes or isolate audio clips for transcription.
- AI summaries: Generate transcript summaries using your own API key, and customize the assistant's prompts for the responses you need.
- Exports: Save subtitles or Markdown, or send results to Notion.
The desktop app runs on Windows and macOS and accepts media formats including MP4, MP3, AAC and M4A. Compatible NVIDIA, AMD and Apple Silicon GPUs can accelerate processing. Offline transcription is separate from the summary feature, which requires your own API key.
