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

# Claude Code Runtime

> Code-specialized runtime powered by Claude Code SDK for development workflows.

Code-specialized execution engine optimized for development workflows. Built on Anthropic's [Claude Code SDK](https://docs.anthropic.com) with advanced file operations and extended history.

***

## Key Features

<CardGroup cols={3}>
  <Card title="Code-Optimized" icon="code">
    Multi-file awareness, repository analysis, safe refactoring
  </Card>

  <Card title="Extended History" icon="clock">
    200 messages with session resumption
  </Card>

  <Card title="Claude Models" icon="robot">
    Opus, Sonnet 4, Sonnet, Haiku
  </Card>
</CardGroup>

***

## Capabilities

| Feature         | Support       |
| --------------- | ------------- |
| Streaming       | ✅ Yes         |
| Tool Calling    | ✅ Yes (MCP)   |
| MCP Servers     | ✅ First-class |
| History         | 200 messages  |
| Session Resume  | ✅ Yes         |
| File Operations | ✅ Advanced    |
| Code Analysis   | ✅ Yes         |

***

## Supported Models

Claude Code Runtime supports **all Claude (Anthropic) models** through LiteLLM. While optimized for Claude models, it can route to any model your LiteLLM proxy is configured to support.

**Commonly used Claude models:**

* **Claude 3.5 Sonnet** (Recommended) - Best balance of capability, speed, and cost
* **Claude 3 Opus** - Maximum intelligence for complex tasks
* **Claude 3 Sonnet** - Balanced performance
* **Claude 3 Haiku** - Fast and economical
* **Any other Claude model** - Including future releases

**Note**: Model support is dynamic and configured at the LiteLLM proxy level, not hardcoded in the runtime.

Learn more: [Anthropic Models](https://docs.anthropic.com/claude/docs/models-overview)

***

## When to Use

**Best for:**

* Code generation and refactoring
* Repository analysis
* Technical documentation
* Long development sessions
* Claude-committed organizations

**Consider alternatives:**

* Multi-provider flexibility → Use [Agno Runtime](/core-concepts/runtimes/agno-runtime)
* General-purpose tasks → Use [Agno Runtime](/core-concepts/runtimes/agno-runtime)

***

## Trade-offs

**Pros:**

* ✅ Code-specialized capabilities
* ✅ 200-message extended history
* ✅ Session resumption
* ✅ Multi-file awareness

**Cons:**

* ❌ Optimized for Claude (not multi-provider like Agno)
* ❌ Slower startup (\~150ms vs \~50ms)
* ❌ Not ideal for general tasks

***

## Framework Links

<CardGroup cols={2}>
  <Card title="Anthropic Docs" icon="book" href="https://docs.anthropic.com">
    Official Claude documentation
  </Card>

  <Card title="MCP Protocol" icon="plug" href="https://modelcontextprotocol.io">
    Model Context Protocol specs
  </Card>
</CardGroup>

***

## Next Steps

<CardGroup cols={2}>
  <Card title="Compare Runtimes" icon="scale-balanced" href="/core-concepts/runtimes/comparison">
    See side-by-side comparison
  </Card>

  <Card title="Agent Setup" icon="robot" href="/core-concepts/agents">
    Configure agents with Claude Code
  </Card>
</CardGroup>