Audiobook to Text: free Whisper transcriber, TXT and SRT

Where it runs

Works in a current Chrome, Edge, Firefox or Safari, on Windows, Mac, Linux and Chromebook computers, Android phones, iPhones and iPads. Nothing to install, no account.

The first transcript downloads a Whisper speech model once (about 43 MB for the quick English one), faster with a graphics card (WebGPU).

4 simple steps.

Drop your audio

An audiobook, a chapter, a narration take or a video: MP3, M4B, WAV, FLAC and more.

Pick a model

Quick for English, Balanced for most books, Accurate when you have a fast graphics card. It downloads once.

Watch it listen

The transcript appears as it goes, with timestamps you can click to hear that moment.

Export

Plain text, text with timestamps, SRT or VTT subtitles.

How it works

The file is decoded on your device in windows of about a minute and a half, each ending at a quiet moment so no word is cut in two. Whisper transcribes each window in a background worker, on your graphics card with WebGPU or on your processor, and the text appears as it goes. The model is downloaded from Hugging Face the first time and stays in your browser’s cache, so later transcripts start straight away and work offline.

Model	Languages	Download
Quick (Whisper tiny.en)	English	about 43 MB (120 MB with WebGPU)
Balanced (Whisper base)	99 languages	about 79 MB (210 MB with WebGPU)
Accurate (Whisper small)	99 languages	about 250 MB (590 MB with WebGPU)

Questions, answered.

Is my audio uploaded to a transcription service?

No. The speech model (OpenAI’s Whisper, MIT-licensed) runs in your browser. It downloads once from Hugging Face and is kept in your browser’s cache; your audio never leaves your device.

How long does it take?

With WebGPU (current Chrome and Edge on most computers) Whisper runs several times faster than real time. Without it, it runs on your processor, which can be slower than real time for the bigger models: try Quick for long English books, or transcribe a chapter at a time.

Which model should I choose?

Quick (tiny, English only) is fast and fine for clear narration. Balanced (base) handles most books and other languages. Accurate (small) is best for names, accents and noisy recordings, and needs WebGPU to be quick.

How do I convert an audiobook to text?

Drop the MP3, M4B or M4A here, pick a model and press Transcribe. Whisper listens to the book in your browser and writes a transcript with a timestamp on every line; search it, then download it as plain text (TXT) or as captions (SRT, VTT).

Can I transcribe Audible books?

Only DRM-free files you have the right to use. Protected Audible files (AAX) are declined. If you want the text of a book you own, its ebook edition is the reliable source.

What can I use a transcript for?

Checking a narration against the manuscript, finding a passage for a pickup, writing chapter titles, making captions for an audiogram or a video, or searching a long book.