Gemini Live API
A real-time multimodal API provided by Google that supports streaming voice and video interaction. It can be combined with function-calling capabilities to build AI assistant applications with real-time perception and action abilities.
Timeline (last 90 days)
The Gemini Live API supports two-way voice conversation, allowing apps to stream microphone audio and camera frames and receive low-latency voice replies in real time, with the session staying open until the user actively closes it.
The above capabilities are now available in the Gemini Live API for developers to use directly.
The Gemini Live API update includes features such as asynchronous function calling, proactive audio, context injection, and background reasoning
The proactive audio feature allows developers and users to prompt the agent to respond only when relevant or when addressed by name
Async function calling allows long-running tools to run in the background while the agent continues conversing with the user
These new features are now available for developers to use
Context injection injects client content into the conversation via send client content without forcing a conversation turn
Google updated the Gemini Live API, for the first time bringing frontier-level reasoning into the native audio interaction experience
Google demonstrated in a technical demo video the use of the Gemini Live API with function calling to build a real-time cooking assistant app called Friendly Meals via Firebase AI Logic.
Through the Gemini Live API, platforms such as Agora, Fishjam, LiveKit, Pipecat, Vercel, and Vision Agents can help developers build voice interfaces.
All Facts (10)
The proactive audio feature allows developers and users to prompt the agent to respond only when relevant or when addressed by name
60%UnverifiedAsync function calling allows long-running tools to run in the background while the agent continues conversing with the user
60%UnverifiedContext injection injects client content into the conversation via send client content without forcing a conversation turn
60%UnverifiedThe Gemini Live API supports two-way voice conversation, allowing apps to stream microphone audio and camera frames and receive low-latency voice replies in real time, with the session staying open until the user actively closes it.
50%UnverifiedThrough the Gemini Live API, platforms such as Agora, Fishjam, LiveKit, Pipecat, Vercel, and Vision Agents can help developers build voice interfaces.
50%UnverifiedThe above capabilities are now available in the Gemini Live API for developers to use directly.
70%UnverifiedThe Gemini Live API update includes features such as asynchronous function calling, proactive audio, context injection, and background reasoning
60%UnverifiedGoogle updated the Gemini Live API, for the first time bringing frontier-level reasoning into the native audio interaction experience
50%UnverifiedThese new features are now available for developers to use
50%UnverifiedGoogle demonstrated in a technical demo video the use of the Gemini Live API with function calling to build a real-time cooking assistant app called Friendly Meals via Firebase AI Logic.
50%