Trascrivi
Turns voice messages and audio files into text and, with your own prompt, into summaries, tasks or translations. Audio is never stored.
What it does
- Transcription from files, microphone, drag & drop or sharing
- Several recordings in a row
- Reusable prompts: summaries, tasks, translations
- Audio never stored: not on disk, not in logs
- Invitation-only accounts with 2FA and recovery codes
- Shared or personal APIs, per user
- Installable as an app, with voice message sharing on Android
Trascrivi started from a simple need: nobody enjoys listening to long voice messages. Upload them, share them straight from WhatsApp or record them, and get the text in seconds.
Privacy first
Audio passes through the server only for as long as needed and is never written anywhere: the database only holds accounts, settings and prompts. Users with their own APIs send audio straight from the browser to their provider, never through my server.
From text to work done
A switch hands the transcript to a language model with your saved instructions: "extract the tasks", "summarise in three points", "translate into English". It works with any OpenAI-compatible endpoint, including self-hosted ones.