Around 98% on a clean recording. The two things that hurt it most are people talking over each other and a microphone far from the speaker — not the format or the file size.

Do I need an account?

Not to try it. To keep the transcript, edit it and download it, yes.

How long can the recording be?

Up to 10 GB per file, which is several hours of audio. Long files take longer, and you see the progress while it works.

What happens to my recording?

It's stored on European Union servers with a retention limit you control, and it's never used to train anything.

How long does an hour-long recording take?

A few minutes. You don't have to sit and watch: we email you when it's done, unless you're still on the progress screen.

Does the audio have to be studio quality?

No. A phone recording on the table works fine if you can understand it by ear. What does cost accuracy is heavy echo, music over the voices, or everyone talking at once.

Try it without an account

Turn any audio into text

Drop a recording and get the transcript with timestamps, the speakers separated and a summary. Start without an account, nothing to install.

Drag your file here

MP3, WAV, M4A, MP4, MOV… up to 10 GB

Speakers separated and named from what's said
Timestamps you can click to jump to the audio
Summary with key points, decisions and tasks

MP3
WAV
M4A
AAC
OGG
FLAC
WMA
AIFF

What this tool does

An hour of audio takes about four hours to type out by hand. It takes us around three minutes, and what comes back isn't a wall of text: it's split by speaker, timestamped, searchable and editable in the browser.

Any format your recorder produced — MP3, WAV, M4A, AAC, OGG, FLAC — up to 10 GB per file. The language is detected on its own, so there's nothing to configure unless you want to.

What we handle that others don't

More than 90 languages, detected on their own. You don't have to tell us which one.
Speaker separation that also works out their names from the greetings in the recording.
Search inside every transcript you own, not just this one.
Your files live on European Union servers and delete themselves when you say.

How it works

1
Drop the audio
Or pick it from your computer. Up to 10 GB, as many files as you want in one go.
2
We transcribe it
Roughly 5% of the recording's length: a one-hour meeting is done in about three minutes.
3
Read, fix, export
Correct anything by clicking on it, then download as DOCX, PDF, SRT, CSV or JSON.

What people use it for

Interviews

Question and answer split by voice, with the timecode to go back and check the quote.

Meetings

The summary lists decisions and tasks, so nobody has to write the minutes.

Podcasts

The transcript is the show notes, the blog post and the subtitles, all at once.

In detail

What makes a transcript come out well or badly

Three things, and none of them is the accent. First is how far away the microphone was: a phone in the middle of a table picks up far more room noise than the same phone in front of the person talking, and that noise is what turns into invented words. Second is whether people talk over each other — when two voices land at once, no system, and no human, gets both.

Third is your own vocabulary. Company names, surnames and industry acronyms are what gets missed most, because they're exactly what appears nowhere except in your recording. Two minutes with the text editor fixes them, and that's where accuracy is really won — not by switching providers.

What you can do with the text afterwards

Searching inside it comes first and gets used most: finding which minute a topic came up without listening to anything again. With the speakers separated you can also filter by who's talking, which in a one-hour interview halves the work.

Then there's getting it out of here: a document to share, an SRT to sit under the video, a spreadsheet to analyse how long each participant spoke. None of that means transcribing again — you export as many times as you like, in whichever format you need.

Questions

And if what you have is the video?

Upload it and you'll get the transcript with timestamps, the speakers separated by name and a summary with the key points. You can start without an account.

Transcribe my recording See prices

Related tools

Worth reading

What is an SRT fileWhat's inside a .srt, why it's still the default subtitle format after thirty years, and the four mistakes that stop one from loading.