The voice/multimodal API is still experimental and heavily rate limited. Raised this 3 months ago and there isn’t change since, do you plan on letting developers build with it at any point in the near future? Just doesn’t give me any confidence to use your products when things are launched and indefinitely on experimental mode.
Hello @batman2212,
Sorry for the late response. The 3 concurrent sessions limit you initially saw is specific to the for the Google AI Free tier .
For higher concurrent sessions, you would need to move to the the Pay-as-you-go tier . You can get 50 concurrent sessions for Tier 1 and 1000 concurrent session for Tier 2.
You will also need to implement the Enable context window compression or Configure session resumption features in your application to handle long-running audio sessions.