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

# Quickstarts

> First-run guide for Direct Mode SDK quickstarts, platform integration first runs, token endpoints, and demo selection.

Quickstarts are the shortest path from credentials to a visible speaking avatar. This section includes Direct Mode SDK quickstarts and the LiveKit Agents Integration quickstart.

Use a Direct Mode quickstart when your app connects the client SDK directly to Motion Server with a Session Token. Use the LiveKit Agents quickstart when you want the fastest packaged Web voice-agent path with LiveKit Agents, Gemini Live, `livekit-plugins-spatius`, and the Web RTC Adapter. For Agora Convo AI setup, see [Agora Convo AI Integration](/agora-convoai/overview).

## Fastest first demo by path

| Path                           | Fastest first-run target                                                                               | Why start there                                                                                                   |
| ------------------------------ | ------------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------- |
| **Direct Mode**                | [Web SDK quickstart](/quickstarts/web-sdk)                                                             | Lowest setup surface: App ID, Avatar ID, Session Token, sample PCM audio, and browser rendering.                  |
| **LiveKit Agents Integration** | [LiveKit Agents quickstart](/quickstarts/livekit-agents)                                               | Packaged Web voice-agent path with LiveKit room setup, Gemini Live, `livekit-plugins-spatius`, and RTC rendering. |
| **Agora Convo AI Integration** | [Convo AI Agent setup](/agora-convoai/convo-ai-agent) plus [Agora client setup](/agora-convoai/client) | Best when you already use Agora Convo AI or TEN Framework; no packaged quickstart lives in this section yet.      |
| **Backend Mode**               | [Backend Mode demos](/resources/demo-projects#recommended-starting-points)                             | Best after you understand the Server SDK and transport responsibilities; it is not a smallest-first quickstart.   |

For a full effort comparison, see [Choose Your Integration Path](/getting-started/how-to-integrate#compare-integration-effort).

## Direct Mode and token servers

Direct Mode clients connect to Motion Server themselves. They still need a short-lived **Session Token**, but that token must be issued by backend code because issuing it requires your `SPATIUS_API_KEY`.

That backend code is only a **token endpoint**. It is not a Backend Mode runtime server: it does not run ASR, LLM, or TTS; it does not connect to Motion Server; and it does not relay audio or motion data. The Direct Mode demos include token-server examples only to show how to exchange your server-side API Key for a client-safe Session Token.

| Server-shaped code | Used by      | Responsibility                                                                                                                           |
| ------------------ | ------------ | ---------------------------------------------------------------------------------------------------------------------------------------- |
| **Token endpoint** | Direct Mode  | Mint short-lived Session Tokens for clients.                                                                                             |
| **Runtime server** | Backend Mode | Own ASR / LLM / TTS, connect to Motion Server with the Server SDK, and deliver encoded audio payloads + motion data payloads to clients. |

## Direct Mode SDK quickstarts

| Platform    | Quickstart                              | What it proves                                                                                                   |
| ----------- | --------------------------------------- | ---------------------------------------------------------------------------------------------------------------- |
| **Web**     | [Web SDK](/quickstarts/web-sdk)         | App ID, Avatar ID, Session Token, sample PCM audio, optional browser realtime conversation, and local rendering. |
| **iOS**     | [iOS SDK](/quickstarts/ios-sdk)         | Native iOS AvatarKit setup with runtime Session Token.                                                           |
| **Android** | [Android SDK](/quickstarts/android-sdk) | Native Android AvatarKit setup with runtime Session Token.                                                       |
| **Flutter** | [Flutter SDK](/quickstarts/flutter-sdk) | Flutter AvatarKit setup on iOS or Android with runtime Session Token.                                            |

## Platform integration quickstarts

| Platform integration | Quickstart                                    | What it proves                                                                                                          |
| -------------------- | --------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------- |
| **LiveKit Agents**   | [LiveKit Agents](/quickstarts/livekit-agents) | LiveKit room setup, Gemini Live voice agent, `livekit-plugins-spatius`, Web RTC Adapter rendering, and microphone flow. |

## Next steps

<CardGroup cols={2}>
  <Card title="Browse all demos" icon="grid" href="/resources/demo-projects">
    Every runnable demo by platform and integration path.
  </Card>

  <Card title="Choose your integration path" icon="route" href="/getting-started/how-to-integrate">
    Compare latency, production ownership, and developer effort.
  </Card>
</CardGroup>
