01
Add the recording
Upload the file or paste the link.
Turn an MP3 audio file into text: a word-timed transcript, .srt and .vtt subtitle files and plain text. There's no picture in an audio file; recording length and the number of speakers decide transcript quality.
Upload the file or paste the link.
Auto-detect is available; setting the language gives more accurate results.
Transcript, .srt, .vtt and plain text download together.
MP3 is a lossy format. In files encoded at very low bitrates (32 kbps and below) high frequencies get crushed and sounds like “s”, “sh” and “f” blur together; that's where most transcript errors come from. If you also have a WAV or M4A, use that instead.
Because there's no picture, processing is quick: the file goes straight to transcription with no download or video decoding step.
The biggest factor is the microphone. A clean audio recording is transcribed almost perfectly; background music, echoey rooms and people talking over each other raise the error rate. Names and brand names are spelled by ear, so it's worth reviewing the output once.
For long recordings the whole file isn't downloaded; only the audio is fetched first, so even hour-long recordings start quickly.
You can pull clips from the same recording right away: the transcript is already word-timed, so clip boundaries never land mid-sentence. No need to re-upload to burn in captions or translate.
With the text in hand, the recording itself becomes searchable: turning it into a blog post, a newsletter summary or show notes takes far longer without a transcript.
The source file is kept only for a while so you can re-process it, then deleted automatically. For meeting and customer recordings that matters: nothing is used to train models. Your outputs stay in your account until you delete them.
You get a transcript and an .srt file. Burning captions into a video needs a picture; combine the audio with a video and use the auto captions tool.
Typical audio files, long podcast episodes included, upload without trouble; the real limit is your plan's monthly minutes, not file size.
Recordings up to three hours are processed, within your plan's monthly minutes.
The transcript is word-timed; speaker labeling isn't available yet.
The source file is kept only for a while so you can re-process it, then deleted automatically. Outputs stay in your account until you delete them.
Burn word-by-word captions into the whole video, with .srt and .txt files.
OpenVideo translatorTranscribe the speech, translate it and burn the translated captions into the video. .srt in both languages.
OpenCrop and resizeTurn a horizontal video vertical or square, following the speaker. Framing only, no captions.
OpenTrim videoSet a start and end and keep that part. Change the aspect ratio too if you like.
OpenSplit videoSplit a long video into equal parts, each labeled 'Part 1/5'.
OpenAdd text and imagesAdd text and/or a logo on top of your video. Choose position, size and style.
OpenProcess 30 minutes of video free every month and see the results. Upgrade if you like it.
No credit card · Cancel anytime