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

# Introduction

> Explore our guides and examples to start running AI models with ModelRunner

## ModelRunner overview

ModelRunner is a platform for running models with a seamless experimentation experience. Each model page provides a hands-on playground (inputs and output preview), a concise model description, related links, and clear cost information—all in one place.

<Columns cols={2}>
  <Card title="Experiment model in the playground" icon="pen-to-square" href="https://modelrunner.ai/bytedance/sdxl-lightning-4step">
    Configure inputs, run generations, and iterate quickly—directly in your browser.
  </Card>

  <Card title="Use the unified API" icon="terminal" href="/docs/api-reference/introduction">
    Call any supported model through one consistent interface. JavaScript SDK available.
  </Card>

  <Card title="Connect via MCP" icon="plug" href="/docs/guides/mcp-server">
    Run every model from Claude, Cursor, or any AI assistant — one OAuth sign-in, no API keys pasted into config.
  </Card>

  <Card title="Browse models & pricing" icon="table" href="https://modelrunner.ai/models">
    The full catalog with per-model pay-per-use prices, by capability — from text-to-image to image-to-3D.
  </Card>
</Columns>

### What you’ll find on each model page

* Input form with typed parameters
* Live output display for quick verification
* Model summary and usage notes
* Related resources and example links
* Pricing and cost guidance