September 18, 2025

The following Press or Say Next Gen Voice enhancements are available:

• Text-to-Speech (TTS) Language andTime to First Byte (TTFB) Key Performance Indicators (KPIs) are now automatically created for you.

• If the Advanced Configuration toggle is enabled and you're using TTS functionality with an ElevenLabs voice, you can now include a Split Long Text configuration.

  • When this configuration is included, TTS messages will be "chunked" (aka split into small messages) so they can be sent for processing more quickly, reducing latency.

• If the Advanced Configuration toggle is enabled and you're using ASR functionality, the language detected throughout the interaction is no longer hard-coded. Language is now detected based on the language spoken by the user.

The following bugs were resolved:

• Previously, when the Press or Say action was configured with Next Gen Voice enabled using the Google-GSRv2 Vendor, an error caused connections to unexpectedly end when the Streaming Speech Start Timeout was exceeded, without any error messages. This issue was resolved, and the system now behaves as expected.

• The Session_ID field for Next Gen Voice default KPIs now displays as expected. Previously, an error caused the word "session" to appear instead of the actual expected ID.