Audio to text is for any recording you'd rather read than replay: a long voice note, a lecture, a job interview, a podcast episode, a doctor's explanation you recorded with permission, a team call. Drop the file and get a transcript split into phrases, each with its timestamp.
It takes the usual audio formats, such as MP3, M4A, WAV, OGG, OPUS, FLAC, AAC and AMR, and video files like MP4, MOV, WebM and MKV, so you don't need to pull the sound out of a video first. Each recording can be up to 120 minutes and 95 MB. The language is detected automatically, and recordings where people switch languages work too.
The transcript opens with a player next to it. Click a phrase to jump to that moment, search for a word, and rename the speaker labels from Speaker 1 to real names. The Summary tab pulls out the key points and the action items, with who does what and by when, and on request writes a follow-up email for the team. You can download the transcript as TXT, Word (.docx), SRT or VTT.
How long you wait depends on the length: a one-minute voice note is ready in a few seconds, a 12-minute call in about half a minute, an hour of audio in about 3 minutes, plus the time it takes to upload.
About accuracy. A clear recording with one or two people transcribes well. Loud background noise, music, people talking over each other or a phone left far away lower the quality. Check names, numbers and technical terms before you quote them. A track with music and no speech may come back with words that weren't said, so it is not the tool for song lyrics.
The file is processed on our server and deleted from there after 24 hours. The transcript is saved in your library, inside this browser. From there you can translate it, make it shorter, or turn it into subtitles for a video.