Frequently asked questions

Yes. It is free, and there is no account or login, ever. There is no daily or monthly quota and no paywall; ads keep the website free.

Your file is transcribed in your browser on your own device, so we set no size or length limit. Very large or long files work too: they take longer and need enough free memory, and the page warns you when your device may be slow.

On clear speech in a supported language it is usually very accurate, but no speech recognition is perfect. Accuracy depends on the audio quality, accents, background noise, overlapping speakers, music, and rare names or jargon. Every line has a confidence score, and low-confidence lines are marked so you can check them first. Always verify a transcript before relying on it for anything important.

Why can results differ between runs, quality settings or other tools? Speech recognition is probabilistic: it chooses the most likely words, and in noisy or ambiguous passages several answers are close. A different quality setting, language setting, vocabulary hint or audio clean-up can change the text, and so can another tool that uses a different system. Speaker names or labels are not added.

• Choose the spoken language instead of auto-detect.
• Add a vocabulary hint: names, brands and jargon that appear in the audio, spelled the way you want them.
• Pick High accuracy or Best accuracy for difficult audio.
• Re-transcribe a finished transcript with another quality or language; your current transcript is kept.
• Start from good audio: less background noise, a close microphone, and one person speaking at a time. Keep "Enhance audio" on.
• Check the marked lines and fix them in the transcript editor.

More than 50 languages, including English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Arabic, Hindi, Chinese, Japanese and Korean. Leave the language on auto-detect or choose it yourself. You can also translate speech into English.

File types. Video: MP4, MKV, MOV, AVI, WEBM, WMV, FLV, M4V, 3GP, TS and MTS. Audio: MP3, WAV, M4A, AAC, FLAC, OGG and OPUS. You can also record straight from your microphone.

Limits. We set no size or length limit: your file is processed on your device, so very large or long files simply take longer and need enough free memory. On a slower device the page tells you before you start and estimates the time.

Nobody else's files are ahead of yours. If you add several files, they are transcribed one after another in your tab, so keep the tab open until they are done.

It depends on your device, the quality you pick and the length of the file. With a capable graphics chip (WebGPU) it is fastest; otherwise it runs on the processor (WebAssembly), which is slower. Fastest is quick; Best accuracy takes longest but makes the fewest mistakes. Progress and the time left are shown live.

The first time you use a quality level, your browser downloads its speech model once and keeps it for next time, so later files start sooner.

Your file never leaves your device. It is transcribed in your browser on this device and is never uploaded to us or to anyone else. Transcripts, your edits and the file are kept in this browser's storage only and deleted automatically after 30 minutes of inactivity, or right away when you press Delete my data now in the Transcripts menu. There is no account. Export what you need before you leave.

The first time you use a quality level, your browser downloads the speech model from Hugging Face and the engine's program files from jsDelivr, and keeps them for next time; nothing of your file is sent with them. When a transcription finishes, your browser adds one to the anonymous counter in the page header, and nothing else.

Ads appear only on information pages and beside finished transcripts, and are served by Google, which uses cookies. No ad is shown while a file is being transcribed, and this website sets no cookies of its own. Details are in the Privacy policy.

Yes. Use the pencil next to a line to change its text or delete the line, or search and replace across the whole transcript. Your edits are included in every export.