Meta-Me Transcribe
Audio in, clean text out, and a voice to read it back. Speak a draft, tidy it, hear it — all in one document, all on your own hardware.
What it does
- Record straight into the browser or upload a file — Whisper handles the transcription
- Raw transcripts are tidied into clean Markdown — headings, paragraphs, punctuation
- Hear it back — synthesise the whole document or a single section in a natural voice
- Headings split a document into addressable segments — re-record or re-voice just one
- Speech recognition and synthesis run on our own machines — no third-party speech cloud
- Part of Meta-Me — one account, your data, no ads, no tracking
Your voice doesn't go to a speech cloud
Transcription (Whisper) and synthesis (Kokoro) run on machines we own, not on a third-party speech API. Your recordings are stored as files on our own volume, isolated per user by row-level security. The only external model we call is the one that tidies a raw transcript into Markdown.
Self-hosted Whisper Self-hosted Kokoro TTS Row-level security No ads, no trackers
Part of Meta-Me
One subscription. 7 tools.
Your subscription unlocks 7 more apps, all for the same £8/month. Start with a 30-day free trial.