Word-level timestamp transcription in JSON
Transcribe audio or video with AI and get a JSON file with the timing of every word, plus its translation into 53 languages. Built for editors, developers and automated workflows.
Restricted content
Sexually explicit, pornographic or similar content is not translated. If you upload this kind of material, the translation will be rejected and not processed. Please only use appropriate content (movies, series, podcasts, documentaries, etc.).
Select your file
Drag audio or video here
or click to choose (MP3, WAV, MOV, MP4, etc.)
You get 5 free minutes of audio per week.
Per-word timestamps
Besides the text of each line, the JSON includes when every word starts and ends. Use it to build your own animated subtitles, sync text with video in your app or search inside a recording.
Original and translation
You get two files: the transcript in the original language of the audio and the translation into the language you pick out of 53.
From any audio or video
Works with MP3, WAV, M4A, AAC, FLAC, OGG, WMA, MP4, MOV, MKV, WEBM, AVI, FLV, WMV and M4V.
How it works
- Upload your audio or video file.
- Pick the language of the audio and the language you want to translate into.
- Pick the output format and, if the format supports it, the color and karaoke mode.
- Download the transcript in the original language and the translation from My files. We email you when they are ready.
Frequently asked questions
Which engine does the transcription?
Whisper, the speech recognition model. The translation is done by an AI language model.
What happens to my file?
The original audio is deleted as soon as processing finishes. Only the generated subtitles are kept so you can download them.
Is it free?
You get 5 free minutes of audio every week. For longer files you can pay for just that file (from USD 2) or pick a monthly plan. You need an account to upload files.