Guide

Quickstart: Build your first voice agent

Get a voice AI agent running in under 10 minutes. Uses Python and the Agents SDK with a hosted LLM and STT.

Read guide

Video

Deep dive: WebRTC architecture explained

20-minute walkthrough of how the media transport layer works and how to optimize for low latency applications.

Watch video

Docs

API reference: Agents SDK

Complete API reference for the Agents framework — events, lifecycle hooks, STT/TTS adapters, and session management.

Open docs

Tutorial

Deploy your agent to production

Step-by-step tutorial covering containerization, environment config, scaling workers, and monitoring in production.

Follow tutorial

What's new

Agents SDK v1.0 release notes

Stable API, improved pipeline performance, new STT adapters for Whisper and Deepgram, and breaking changes from beta.

See what's new

Article

How we reduced voice latency to under 300ms

Engineering deep-dive into the pipeline optimizations, edge deployment, and adaptive jitter buffer that made it possible.

Read article