Your files stay on your device

Audio to text converter

Transcribe files or microphone recordings on your device with a local Whisper model, then edit lines and export TXT, SRT, VTT, JSON, or MD.

Pick a file to decode it and draft the transcript.

What leaves your browser

When the browser recognizes speech, your voice goes to the browser maker's speech service, which this page never sees; the decoding, splitting, editing, and exports all stay in this tab. This page leaves your browser only when you press the button; never on load, never in the background. Nothing is sent anywhere else. Google Analytics still measures the visit to this page, and it never receives what you run through this tool.

This tool does not run a Whisper model and does not take podcast links or RSS feeds; recognition is the browser's own speech service where it exists, and without it you still get timestamped segment lines to fill in by hand. Uploads stay minutes-long takes, not 500 MB drops, and nothing is stored after you leave the page. Songs and lyrics transcribe poorly; the segmenter follows loudness, so a paused, quiet take needs the silence gate lowered.
Options

More audio tools

- Convert audio (/audios/convert)
- Trim audio (/audios/trim)
- Normalize audio (/audios/normalize)
- Bass guitar tuner (/audios/bass-guitar-tuner)
All audio tools (/audios)

Was this tool useful?
YesNo

Canonical HTML: https://nutter.tools/audios/audio-to-text
Markdown version: https://nutter.tools/audios/audio-to-text.md
Plain-text version: https://nutter.tools/audios/audio-to-text.txt
Agent index: https://nutter.tools/llms.txt