Skip to main content

A powerful TypeScript SDK built on AI SDK for creating real-time voice and video AI agents with streaming text generation, parallel TTS processing, and intelligent conversation management.

Quick Start

Get up and running with Voice Agent in minutes.

Installation

Install the SDK and set up your project

Quickstart Guide

Build your first voice agent in under 5 minutes

Key Features

Everything you need to build production-ready voice AI applications.

Streaming TTS

Chunked streaming TTS with parallel generation for low latency

Barge-in Support

Natural conversation flow with intelligent interruption handling

Memory Management

Configurable history limits and sliding-window memory

WebSocket Protocol

Full-featured real-time communication protocol

Agent Types

Choose the right agent for your use case.

VoiceAgent

Audio transcription, streaming LLM, and TTS generation

VideoAgent

Vision-enabled models with video frame and audio processing

Explore the SDK

Core Concepts

Learn about the architecture and design patterns

API Reference

Complete API documentation for all agents and managers

Examples

Real-world usage examples and integration patterns

Build docs developers (and LLMs) love