Tools / Audio to text
Transcribe audio to text for free
Join the teams making the switch to Cartesia
How to convert audio to text
Drop in an MP3, WAV, M4A, WebM, FLAC, or OGG file up to 10 minutes long. No account needed.
The file plays while the transcript fills in, so you can read along.
Copy the transcript or save it as a .txt file.
Transcribe voice memos, interviews, and calls
Voice memos
Voice memos from an iPhone are M4A files, so you can drop one in without converting it.
Interviews
A new paragraph starts each time a speaker finishes a turn, so questions and answers don't run together.
Voicemails and calls
Phone numbers, dates, and email addresses in a voicemail or call are transcribed correctly, so you can copy them straight from the transcript.
Recordings in other languages
Spanish, French, Hindi, and Japanese recordings work too, and you don't need to pick the language first.
Explore more voice tools
Audio to text FAQs
How do I convert an audio file to text?
Drop the file on the upload box at the top of the page, or choose it from your device. The audio plays while the transcript fills in. When playback finishes, copy the transcript or download it as a .txt file. Files can be up to 10 minutes long and 200 MB.
Is Cartesia's audio to text converter free?
Yes. Drop in a file and get the transcript without an account.
Which file formats can I convert to text?
MP3, WAV, M4A, WebM, FLAC, and OGG. If your file is in another format, convert it to MP3 first.
How do I transcribe a voice memo?
iPhone voice memos are M4A files, and Cartesia's audio to text converter accepts them as they are. In Voice Memos, tap Share, then Save to Files. Then tap the converter's upload box and choose the memo from Files. On a computer, drag the memo onto the upload box.
What languages can Cartesia's audio to text converter transcribe?
English, French, Hindi, Japanese, and Spanish. The language is detected automatically, so there is nothing to set before you upload.
How accurate is Cartesia's audio to text converter?
The converter runs on Cartesia's Ink-2 speech-to-text model, ranked #1 for streaming speech-to-text accuracy on the VoiceArena US-English leaderboard and Artificial Analysis' streaming benchmark. Ink-2 transcribes phone numbers, dates, and email addresses accurately, even with background noise.
How do I transcribe a recording longer than 10 minutes?
The free converter handles files up to 10 minutes. For longer ones, use Cartesia's batch speech-to-text API, which takes a complete file and returns one transcript. You need a free Cartesia account and an API key.
Transcribe longer recordings with the API
For files over 10 minutes, Cartesia's batch speech-to-text API takes a complete file and returns one transcript. To transcribe live audio in your own app, stream Ink-2 over a WebSocket and get each word back as it's spoken.