Gemini Live API
A comprehensive API platform for Google's Gemini Live models, providing WebRTC support, HTTP endpoints, and webhook integration for building next-generation conversational AI applications.
Perfect for developers looking to integrate advanced multimodal AI capabilities into their products with enterprise-grade reliability, comprehensive tool management, and seamless real-time communication features.
WebRTC Support
Built-in WebRTC capabilities for real-time audio and video communication with Gemini Live models.
Simple HTTP API
Easy-to-use REST API endpoint for seamless integration with your existing applications and workflows.
Webhook for session events and tool management
Real-time notifications and event-driven architecture with customizable webhook endpoints for session events and tool management.
Natural conversation
Filler, humming, pauses, dictionary hints and natural language handling.
Conversation profiles
Default prompts and configuration for different conversation types out of the box.

Quick Start
Create Session
const rtc = new RTCPeerConnection();
const offer = await rtc.createOffer();
await fetch("https://rti.livetok.io/connection")
Session Termination WebHook
Receive notifications when a session ends with duration and transcription data.
{
metadata: {}
duration: 3223,
transcription: [
]
}
Tool Calling Webhook
Handle tool execution requests from the AI agent during conversations.
{
}
Pricing
Choose the plan that fits your needs
Free
- •2 concurrent calls
- •Bring your own key
- •All Gemini models and voices available
- •Hooks and Tools
Premium
- •Unlimited calls
- •Document indexing (RAG)
- •Recording