Audio To Text

Drop in a recording and get its words back as text you can copy, edit, or download.

Your file never leaves your device
Drop your files here Tap to choose a file

or browse your device — you can also paste (Ctrl+V)

Your camera and gallery open from the file picker

One file at a time. Files over 200 MB may be too large for browser processing.

    Your result will appear here.

    Free, no account, no sign-up. Your file never leaves your device — this page works even offline once loaded.

    Drop in a recording and get its words back as text you can copy, edit, or download. This audio-to-text tool runs a built-in speech-recognition model entirely in your browser: it listens to your clip and writes out a transcript, which appears on the page and downloads as a plain text file. It handles WAV, MP3, M4A, OGG, and FLAC recordings, and works best on clear speech recorded close to the microphone — a voice memo, an interview clip, a lecture excerpt, or a meeting note. Two honest limits shape the tool. It understands English only, because the bundled model is English-only, and each clip can be at most 60 seconds long. It is also slow: the model downloads once on first use, about 41 MB, and in testing four seconds of audio took roughly 25 seconds to transcribe. The transcript is a useful draft, not a certified record — check it against the recording before quoting anything important. Your audio never leaves your device.

    How to use Audio To Text

    1. Choose your file drop it onto the box above or browse your device.
    2. Set the options adjust any settings below the box, if the tool has any.
    3. Click Convert the conversion runs entirely in your browser — nothing is uploaded.
    4. Download the result save the converted file to your device.
    5. Convert another clear the tool and choose a new file — no sign-up, no limits.

    Format facts

    InputA file from your device.
    EngineBuilt-in browser engine
    CategoryAudio Tools

    Frequently asked questions

    How long does transcription take?

    Longer than real time. The speech model downloads once on first use (about 41 MB), and in testing four seconds of audio took roughly 25 seconds to transcribe. A short voice memo finishes in under a minute; a full 60-second clip takes a few minutes. Keep the tab open while it works.

    Which languages can it transcribe?

    English only. The bundled model is trained for English speech, so recordings in other languages will come back wrong or empty. There is a language setting on the page, but it only offers English — other languages need a different model.

    Which audio formats work?

    WAV, MP3, M4A, OGG, and FLAC recordings. Each clip can be at most 60 seconds long — longer recordings should be split into shorter clips and transcribed one at a time.

    How accurate is the transcript?

    Good for clear speech, weaker for the hard cases. The model is the tiny bundled variant, so it handles a single clear speaker well but struggles with heavy background noise, overlapping voices, strong accents, mumbling, and music. Treat the transcript as a draft: read it through and fix names, numbers, and anything you plan to quote.

    Does my recording leave my device?

    No. The whole transcription runs in your browser. Your audio is never uploaded, and the transcript is produced locally and downloaded directly to your device.