Accessing Live Meeting Transcript or Live Audio Feed?

Hello,

I’m helping run a software development class at a university and there are a number of student teams that are interesting in building Zoom applications for their term projects. Many of the proposed projects require access to either the live transcript of a meeting in progress or access to the live audio feed. If we had access to the live transcript that would be perfect, but even getting access to the live audio feed would allow us to implement our own automatic speech recognition system and go from there.

However, I’ve been looking around at the API and SDKs for how to access these kinds of functionalities and it seems like either it doesn’t exist (in the case of accessing the live transcript) or access has been removed fairly recently (in the case of the raw audio buffers/feeds). Of course, it could also be the case that I have missed this piece of functionality in either the API/SDK documentation, in which case I would be thrilled to be pointed in the right direction.

Is there any possibility of us programmatically getting access to the live transcripts of a meeting in progress and being able to use that within our application?

3 Likes

I would really love an answer to this question

1 Like

Following. We would love to be able to access a live transcript text stream to inject into live-streams.

Hey @Marko @akochendorfer @solomon,

Thanks for using the dev forum!
We do not currently offer a way to access the live transcripts in the SDK’s or API’s.

Thanks!
Michael

hey Michael. I thought I’d just check in to see if there has been any changes on this. Would also love to see live transcripts. First a webhook that it was turned on. But ultimately accessing the live transcript in a live streaming manner
thanks!
Gary

Hi @gpinkham,

As of v5.9.0, the SDK does have a callback for when a live transcription message was received. For example, on Android the callback used would be onLiveTranscriptionMsgReceived.

Thanks!

Thank Jon. I have to admit I didn’t notice that I had switched from the Webhook channel to the SDK channel so I was focused on webhooks… but there’s a chance I may do something with the SDK so that will useful.
thanks again!

This is interesting! A few follow up questions @jon.zoom:

  • Does this mean that the implementing client should have access to the live transcription of the entire meeting at the transcription message level?
  • And would this be available via the web SDK?

@jon.zoom Still interested in :point_up: – Is it better to create a distinct question?