Free minutes daily · Files not stored
Audio to Text Converter
Upload a recording and transcribe audio to text with timestamps, in dozens of languages. 10 free minutes a day without an account, 30 with a free Google sign-in.
Ready
Transcribe audio or video
or drop an audio or video file here
Only the sound is sent, in pieces, and we don't store it. Copy or download the transcript before you leave.
Three steps
How to convert audio to text
- 01
Choose a file
Any recording: MP3, M4A, WAV, OGG, FLAC, WebM and MP4, plus MOV in Chrome and Safari. The cost in minutes shows before you start.
- 02
Press Transcribe
Leave the language on Auto-detect, or pick it. Each 10-minute piece is sent and transcribed in turn.
- 03
Copy or download
Read it with timestamps, as paragraphs, or download TXT, SRT or VTT.
Files it takes
Audio and video: MP3, M4A, WAV, OGG, FLAC, WebM and MP4, plus MOV in Chrome and Safari. Only the sound of a video is used. Your browser decodes the file, so anything it can play almost always works. Files run up to 30 minutes when you're signed in.
Accuracy and languages
The transcript comes from an AI speech recognition model that handles dozens of languages and detects which one it hears. Clear speech close to the microphone comes out nearly word for word. Background music, people talking over each other and phone-call audio lower the accuracy, and names are the most common mistake. Read it through before you publish.
What it costs after the free minutes
Free minutes come first: 10 free minutes a day without an account, 30 with a free Google sign-in. Paid credits for more minutes and longer files are coming soon. See pricing. If part of a file fails, the minutes for that part come back.
Your file stays yours
Your browser decodes the file and keeps only the sound, as plain 16 kHz mono audio. That is sent in 10-minute pieces, transcribed, and the text comes back to your screen. We store neither the audio nor the transcript, and we never use them to train anything. Our speech recognition provider doesn't keep it by default, though it may keep logs for up to 30 days to look into errors or abuse. Copy or download the transcript before you leave the page.
Questions
Frequently asked questions
How accurate is it?
Clear speech with little background noise comes out close to word for word. Music, crosstalk, strong accents and poor microphones cause more mistakes, so read the transcript before you publish it. Names and rare terms are the most common errors.
Which files work?
Most audio and video files: MP3, M4A, WAV, OGG, FLAC, WebM and MP4, plus MOV in Chrome and Safari. Your browser does the decoding, so a file your browser can play almost always works. If one doesn't, convert it to MP3 and try again.
How long does it take?
Most of the wait is the upload: about 2 MB per minute of audio, so the speed depends on your connection. Decoding a long file in your browser adds a little time at the start.
Is it free?
Yes, up to a daily limit: 10 free minutes a day without an account, 30 with a free Google sign-in. Free minutes reset at midnight UTC. On a very busy day the site's free capacity can run out. Paid credits for more minutes and longer files are coming soon. See pricing.
Do you keep my file?
No. Your browser pulls the sound out of the file and sends it to us in pieces. We pass each piece to our speech recognition provider and send the text back to you. We store neither the audio nor the transcript. The provider doesn't keep it by default, though it may keep logs for up to 30 days to look into errors or abuse. See the privacy policy.
Does it label who is speaking?
Not yet. The transcript has timestamps but no speaker names, so add those yourself if you need them.
Tools