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

# Moxus AI overview

> Moxus AI is a unified LLM API gateway for accessing multiple model providers through OpenAI, Anthropic, and Google-compatible APIs.

Moxus AI is a unified LLM API gateway. You can use one API key and a small set of compatible endpoints to access models from OpenAI, Anthropic, Google, DeepSeek, Qwen, Zhipu, MiniMax, xAI, and other providers.

<Tip>
  Use one API key with all three compatible protocols. Existing OpenAI SDK integrations can keep their current request flow: set the Base URL to the OpenAI-compatible endpoint and use a Moxus AI API key.

  <CodeGroup>
    ```text Platform theme={null}
    https://moxus.cloud
    ```

    ```text OpenAI-compatible theme={null}
    https://moxus.cloud/v1
    ```

    ```text Anthropic-compatible theme={null}
    https://moxus.cloud
    ```

    ```text Google-compatible theme={null}
    https://moxus.cloud/v1beta
    ```
  </CodeGroup>
</Tip>

<Columns cols={2}>
  <Card title="Make your first call" icon="rocket">
    Create an account, generate an API key, and call a model with cURL or Python.
  </Card>

  <Card title="Explore modules" icon="layers">
    Browse image understanding, image generation, PDF, and Markdown modules.
  </Card>

  <Card title="Connect developer tools" icon="monitor-cog">
    Configure Cursor, Claude Code, Codex, Roo Code, or browser extensions with endpoint and API key.
  </Card>

  <Card title="Call from code" icon="code-xml">
    Send requests from external code through OpenAI, Anthropic, or Google-compatible protocols.
  </Card>
</Columns>

## What Moxus AI solves

Without an API gateway, using multiple providers means managing separate accounts, credentials, protocols, invoices, and operational behavior. Moxus AI centralizes that work: integrate once, choose models by name, track usage in one place, and keep your existing SDK workflow.

## Core capabilities

* Unified API keys for multiple model providers.
* OpenAI, Anthropic, and Google-compatible protocols.
* Usage logs for request status, token usage, and cost tracking.
* Support for text, image, PDF, Markdown, and structured outputs.
* Key-level controls for quota, expiration, model scope, and IP restrictions.

## Next steps

* [Multimodal modules](/en/overview/multimodal)
* [Cursor](/en/integrations/clients/cursor)