> ## Documentation Index
> Fetch the complete documentation index at: https://docs.outspeed.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Outspeed API

> Embed low-latency voice interaction in your apps with Outspeed API

The Outspeed API enables you to build fast, natural voice experiences by connecting to our hosted speech-to-speech stack.

With Outspeed, you can send and receive both text and audio in real time, leverage voice activity detection, call external functions, and more.

<Note>
  Outspeed API is fully compatible with the OpenAI Realtime API.
</Note>

## How It Works

Outspeed API is event-driven. After connecting via WebRTC, your app sends [Client Events](/api-spec/client)
and listens for [Server Events](/api-spec/server) to drive the conversation.

Getting started is easy:

* Install the [JS SDK](https://www.npmjs.com/package/@outspeed/client)
* Install the [React SDK](https://www.npmjs.com/package/@outspeed/react)
* Install the [Swift SDK](https://github.com/outspeed-ai/OutspeedSwift)

Each SDK manages the WebRTC connection and event flow for you, so you can focus on building your app.

## Key Capabilities

* **Text & Audio**: Send and receive both text and audio
* **Realtime Responses**: Low-latency replies for natural conversations
* **Function Calling**: Call external tools and services

## API Events

The API is organized around events:

<CardGroup>
  <Card title="Client Events" icon="arrow-up-right-from-square" href="/api-spec/client">
    Events your app can send to the Realtime server.
  </Card>

  <Card title="Server Events" icon="arrow-down-left-from-square" href="/api-spec/server">
    Events your app receives from the Realtime server.
  </Card>
</CardGroup>