Does bitrate affect accuracy?

Very little from 64 kbps upward. Below 32 kbps, combined with a distant microphone, a difference does show.

Mono or stereo, which is better?

It makes no difference. Mono is enough for speech and takes half the space; stereo doesn't improve accuracy.

What size can the file be?

Up to 50 GB, more than enough for any continuous recording.

Wouldn't uploading a WAV be better?

For long-term archiving, yes. For transcribing, the difference in accuracy is minimal and the file is ten times heavier.

Does a lower-quality MP3 transcribe worse?

Less than you'd expect. Below 32 kbps it becomes noticeable, but a typical recorder file at 48 or 64 kbps transcribes just as well as one at 320 kbps. What matters most is microphone distance.

Can I upload several MP3 files at once?

Yes. Upload them together and they'll be processed one after another, each as an independent transcript with its own text and summary.

No account required

Turn an MP3 file into text

The most common audio format, transcribed with timestamps and every voice identified, in a few minutes.

Drag your file in here

MP3, WAV, M4A, MP4, MOV… and any other audio or video, whatever the size

Any bitrate, mono or stereo
Automatic language detection out of more than 90
Text, subtitles and summary in a single process

MP3

About this tool

MP3 is the usual format for recorders, mobile apps and podcast platforms, and it's perfectly suitable for transcribing. Compression removes frequencies the human ear can't perceive, the same ones speech recognition doesn't need either.

From 64 kbps upward, accuracy barely changes with bitrate. What does affect the result is the distance to the microphone and several people talking at once.

Features

Compatible with any bitrate from 32 kbps upward, mono or stereo, fixed or variable.
Long files are processed in full: a three-hour podcast isn't split up.
ID3 tags and the file name have no effect on the result.
Voice identification even in mono mixdowns of several microphones.

How it works

1
Upload the MP3
Up to 50 GB: several hours of recording in a single file.
2
It's transcribed automatically
About three minutes per hour of audio, with progress visible at all times.
3
Download the result
DOCX, PDF, SRT, CSV or JSON, including your corrections.

Use cases

Voice recorders

Most generate MP3 by default. Upload it directly.

Podcast episodes

The published file is enough; you don't need the original recording session files.

Voice notes

Any recording from a phone transcribes just as well.

In detail

Where the MP3 files that reach this page come from

Usually from a voice recorder. These devices record at low bitrates — 32, 48 or 64 kbps — to fit many hours into little storage. This doesn't affect transcription: at that level the voice remains intact, and what's removed are frequencies not used in speech.

The other common source is a video with the audio already extracted, or a downloaded podcast. Both usually arrive at 128 kbps or higher, more than enough. Compressing them further before uploading offers no benefit: minutes are calculated by duration, not by file size.

When the recording runs close to three hours

Three hours of MP3 take up just over 150 MB, so size is rarely the issue. What's worth checking is whether the file begins with an irrelevant section: trimming it beforehand reduces minute usage.

If the same file contains several distinct sessions, it's advisable to split it. Not for technical reasons, but because searching within a three-hour text is inconvenient, and the summary has to cover unrelated topics.

Frequently asked questions

Need to transcribe a video?

Upload it and get the text with timestamps, each voice identified and a summary. No account needed to start.

Transcribe my recording See pricing