Changelog

Mid-session Listen updates, STT latency, session length limit & Google Enterprise Agent API

UpdateListen: switch model and language mid-session

UpdateListen now accepts model and language changes, so you can switch speech-to-text models and languages without restarting the session. Previously the provider identity had to match the current session.

1
{
2
  "type": "UpdateListen",
3
  "listen": {
4
    "provider": {
5
      "type": "deepgram",
6
      "version": "v1",
7
      "model": "nova-3-general",
8
      "language": "es"
9
    }
10
  }
11
}

Send the fields that apply to the model you are switching to: language for V1 (Nova) models, and keyterms, eot_threshold, eager_eot_threshold, eot_timeout_ms, and language_hints for V2 (Flux) models. Keyterms can only be updated mid-session for Flux models — set Nova-3 keyterms in the Settings message at the start of the session. The server responds with a ListenUpdated acknowledgement once the changes are applied.

For details, see the UpdateListen documentation.

STT latency in LatencyReport

LatencyReport now includes stt_latency, the time from audio received to transcript produced. Every turn reports the full STT → LLM → TTS breakdown.

1
{
2
  "type": "LatencyReport",
3
  "stt_latency": 0.12,
4
  "ttt_token_latency": 0.34,
5
  "tts_latency": 0.18,
6
  "total_latency": 0.64
7
}

For details, see the Latency Report documentation.

Maximum session length of 2 hours

Voice Agent sessions now close automatically after 2 hours. Two new codes mark the limit:

KeepAlive does not extend the limit. To continue a conversation past 2 hours, open a new connection and replay the prior turns in agent.context.messages.

For details, see Maximum session length.

Google Enterprise Agent API support

Managed Google LLMs now run on either of Google’s APIs: the AI Studio API or the Gemini Enterprise Agent (GEA) API, formerly the Vertex API. GEA adds regional Google endpoints for models where available. Select the API with think.provider.version from any Deepgram Voice Agent endpoint: ai-studio-v1beta for AI Studio, gemini-enterprise-agent-v1 for GEA. The existing v1beta value remains valid as an alias for ai-studio-v1beta.

1
{
2
  "agent": {
3
    "think": {
4
      "provider": {
5
        "type": "google",
6
        "version": "gemini-enterprise-agent-v1",
7
        "model": "gemini-2.5-flash"
8
      }
9
    }
10
  }
11
}

Defaults when you omit version:

Fewer Gemini models are available on the EU GEA endpoint than on global AI Studio. For current availability by region, see Google’s documentation.

For details, see Regional Endpoints and LLM Models.

Fixed

Resolved a sample rate issue that prevented some users from using the Flux TTS early access model in the Voice Agent.