# What's new in the Gemini Live API

## Executive summary

Google DeepMind introduced significant updates to the Gemini Live API, enhancing real-time voice and vision interactions. Key advancements include asynchronous function calling for background tool execution, proactive audio for relevant agent responses, and context injection via `send client_content`. Most notably, the API now supports frontier-level background reasoning, allowing complex tasks (like generating SVGs) to run while the agent maintains conversational responsiveness.

## Key takeaways

- Asynchronous Function Calling: The API supports async function calling, enabling long-running tools to execute in the background while the agent continues the conversation, improving perceived latency and user experience. (See demonstration of checking an order ID 1234 XYZ)
- Proactive Audio and Context Injection: Proactive audio ensures the agent only speaks when relevant. Context can be seamlessly injected using `send client_content` without forcing a verbal turn, useful for backchanneling information. (Demonstrated early in the video)
- Frontier-Level Background Reasoning: Higher reasoning capabilities are now available in the Live API, allowing the model to handle complex, time-consuming tasks (e.g., drawing an SVG) in the background while remaining highly responsive to ongoing user dialogue.

## Technical details

- Gemini Live API Capabilities: The API supports low-latency, real-time voice and vision interactions using models like Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. (Metadata)
- Context Management: Context can be injected without forcing a turn using the `send client_content` feature, allowing for non-verbal information transfer. (Demonstrated at 120)
- Model Comparison: The demonstration highlighted the difference between the standard native audio model and the specialized 'Max high reasoning model' when performing complex tasks like SVG generation. (Demonstrated at 320)

## Practical implications

- Build engineers can now design conversational AI experiences that handle complex, multi-stage tasks (like background data retrieval or graphic generation) without blocking the user interface or requiring explicit user prompts.
- The combination of proactive audio and context injection allows for highly natural, low-friction user interactions, improving the overall perceived quality of the AI agent.
- The ability to manage background tasks while maintaining conversational flow is critical for building robust, enterprise-grade AI applications.

## Topics

Gemini Live API, Generative AI, Real-Time Voice Interaction, Asynchronous Processing, Context Injection, Gemini Live API overview, Live API in Google AI Studio

Source: https://www.youtube.com/watch?v=3CyW24Pkz4o