GA / Production Release Timeline for gemini-3.1-flash-live-preview on Vertex AI

Gemini API
, , ,
1

Hi Google Gemini / Vertex AI Team,

We are currently using the following Gemini Live model in a production application on Google Cloud Vertex AI:

gemini-live-2.5-flash-native-audio

We are now evaluating the newer Live model:

gemini-3.1-flash-live-preview

According to the current Gemini documentation, gemini-3.1-flash-live-preview is still listed as a Preview model. We understand that it was released in March 2026 and is designed for low-latency, real-time audio-to-audio applications.

Since our application is already running in production, we would prefer to migrate only when the Gemini 3.1 Flash Live model is available as a Generally Available (GA) / production-supported model on Vertex AI.

Could someone from the Google team please clarify:

  1. Is there an expected GA / production release date or timeline for gemini-3.1-flash-live-preview on Vertex AI?
  2. Will the GA model use a production model ID such as gemini-3.1-flash-live, or will the model name be different?
  3. Will the GA release be supported through the normal Vertex AI production endpoint using Google Cloud authentication?
  4. Will it support the same real-time native audio / audio-to-audio capabilities as the current preview?
  5. Will features such as function calling, session resumption, Google Search grounding, and production quotas be supported?
  6. Is there an early-access or allowlist program for customers who want to evaluate the production version before GA?
  7. Is there a recommended migration path from:

gemini-live-2.5-flash-native-audio

to:

gemini-3.1-flash-live-preview

We are particularly interested in this information for production planning, as we do not want to depend on a Preview model for a customer-facing real-time voice application.

Any guidance from the Gemini / Vertex AI product team regarding the expected production availability would be greatly appreciated.

Thank you.