Hi support, I was wondering if there is a way to combine the partial transcripts to display in a chat message. Now the partial transcript contains duplicate text per partial and does not continue from where the previous partial ended. For example: 1st partial: "Hello world" 2nd partial: "Hello world, how are" 3th partial: "are you doing?" What I would expect is a consistent set of partials that I can combine into one final message. For example: 1st partial: "Hello world" 2nd partial ", how are" 3th partial: "you doing?" Now the partials cannot be used to build up a full message. How can I fix this? My second question is related to the final messages for the assistant transcript. The delay before I get a final assistant message is long before the voice has started speaking. I would like a faster final message that's closer to the voice output, so I can construct a final chat message of what the voice assistant is saying. The delay between the voice and the assistant final transcript is now too slow.

It looks like you're experiencing incremental transcriptions where each partial transcript overlaps with the previous one. To manage this, you might need to implement logic in your system to combine these partials effectively into a single coherent message. One approach is to stream the partial transcripts and dynamically remove or format the redundant segments as new partials arrive. Also, ensure that your transcriber configuration includes settings that support interim results for real-time updates, similar to enabling

interim_results

in transcription settings. Let me know if you need more help!