Home/Integrations/Transcriptions/ElevenLabs Scribe
Transcripts with speakers and timings, automatically.
AddPipe extracts an audio track from every recording. Send it to ElevenLabs Scribe from a webhook and you get back a transcript that knows who spoke and exactly when.
Update: Update: The AddPipe Platform now has built-in support for audio transcription, which allows you to automatically detect the language of the audio and transcribe it. You can find more information about this feature in the documentation.
name_audio.mp4Who said what
Speaker diarisation matters the moment a recording has two voices in it, which is most interviews and every conversational study.
Word-level timings
Every word carries a timestamp, so a search result can jump the player to the exact second rather than to the top of a twenty-minute recording.
Audio only, not the video
Sending the extracted audio track rather than the mp4 makes each request smaller, faster and cheaper, with no loss of accuracy.
What this is good for.
Interviews
Two voices, correctly attributed, which is the difference between a usable transcript and a wall of text.
Searchable archives
Word timings let a search result open the recording at the moment the phrase was said.
Summaries
A clean transcript is the input a language model needs to summarise a hundred responses reliably.
Details.
What does Scribe give me that a basic transcript does not?
Speaker diarisation and word-level timings, so you can tell who said what and jump the player to the exact moment a phrase was spoken.
Which file do I send?
The extracted audio track, name_audio.mp4, produced by the transcoding engine alongside the video. Enable it in the Transcoding Engine settings.
Do I need to push to storage first?
Not necessarily. You can fetch the audio from AddPipe when the webhook arrives and post it to the ElevenLabs API. Pushing to your own bucket first is cleaner if you want a permanent copy.
Which webhook event should I use?
video_converted once transcoding is finished, or video_copied_s3 if you are reading the file from your own bucket.
How do I tie the transcript back to the recording?
By the recording name, which the webhook carries. Store the transcript against it and both the video and the text are reachable from one identifier.
Can I use a different transcription service?
Yes. The audio track is a plain file, so Amazon Transcribe and OpenAI Whisper via Replicate work through exactly the same pattern.
Related integrations.
Turn every recording into text.
Webhooks and the audio track output are open during the 14-day trial. No card required.