# RunAPI

> Unified API for AI models, with guides, API reference, SDKs, CLI, and tool integrations for developers.

## Guides

- [Quickstart](https://runapi.ai/docs/guides/quickstart.md): Create an API key, verify your connection, and choose the right RunAPI integration path.
- [Authentication](https://runapi.ai/docs/guides/authentication.md): Create an API key and authenticate requests to RunAPI.
- [Quickstart](https://runapi.ai/docs/guides/task-api/quickstart.md): Create an asynchronous Task and handle polling, completion, failure, and callbacks.
- [Callbacks](https://runapi.ai/docs/guides/task-api/callbacks.md): Securely receive and verify Task callback deliveries.
- [Quickstart](https://runapi.ai/docs/guides/llm-api/quickstart.md): Call RunAPI language models through a supported synchronous or streaming protocol.
- [Quickstart](https://runapi.ai/docs/guides/platform-management/quickstart.md): Create and govern standard API keys and guardrails with a RunAPI management key.

## API Reference

- [OpenAI Chat Completions](https://runapi.ai/docs/api/openai/chat-completions.md): Send a request through RunAPI's OpenAI-compatible Chat Completions operation.
- [Kling Text-to-Video](https://runapi.ai/docs/api/kling/text-to-video.md): Use the text to video endpoint to create an asynchronous Task.
- [Current Account](https://runapi.ai/docs/api/management/current-account.md): Read the account and user context bound to the API key that made the request.

## Developer Resources

- [SDKs](https://runapi.ai/docs/resources/sdks.md): Install a typed RunAPI SDK, authenticate it with an API key, and choose the right Task lifecycle for your application.
- [CLI](https://runapi.ai/docs/resources/cli.md): Install and use every RunAPI CLI command for model actions, Tasks, files, account information, pricing, callbacks, and Harness.
- [MCP Overview](https://runapi.ai/docs/resources/mcp.md): Start with Hosted MCP, or use Local MCP when your client requires a local process.
- [Hosted MCP](https://runapi.ai/docs/resources/mcp/hosted.md): Connect Claude Desktop, Cursor, VS Code, Windsurf, or another MCP client to RunAPI with OAuth.
- [Local MCP](https://runapi.ai/docs/resources/mcp/local.md): Run the RunAPI MCP Server locally with Claude Code, Cursor, VS Code, Windsurf, or another stdio client.
- [CC Switch](https://runapi.ai/docs/resources/tool-integrations/cc-switch.md): Install CC Switch, add a RunAPI Claude Code Provider, choose a model, and verify the connection.

## Catalog

- [Model catalog](https://runapi.ai/models.md)
- [Pricing](https://runapi.ai/pricing.md)