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

# OpenClaw integration

> Configure MaxAPI as a custom model provider in OpenClaw.

[OpenClaw](https://openclaw.ai/) is a self-hosted AI assistant that connects chat channels such as Telegram, WhatsApp, and Discord to AI models and tools.

This guide configures MaxAPI as an OpenAI-compatible provider.

## Prerequisites

Before you begin, you need:

* A MaxAPI API key
* An available model ID
* Node.js 22 or later
* OpenClaw installed locally

<Steps>
  <Step title="Install OpenClaw">
    Install OpenClaw globally with npm, then start its onboarding flow.

    ```bash theme={null}
    npm install -g openclaw
    openclaw onboard
    ```
  </Step>

  <Step title="Store the API key">
    Store your MaxAPI API key in an environment variable.

    ```bash theme={null}
    export MAXAPI_API_KEY="YOUR_API_KEY"
    ```

    <Warning>
      Never commit a real API key to a configuration repository. Use environment variables or a secrets manager in production.
    </Warning>
  </Step>

  <Step title="Add the MaxAPI provider">
    Add `maxapi` under `models.providers` in your OpenClaw configuration. Replace `YOUR_MODEL_ID` with an ID from the [model list](https://api.maxapi.ai/modelmak).

    ```json5 theme={null}
    {
      models: {
        mode: "merge",
        providers: {
          maxapi: {
            baseUrl: "https://api.maxapi.ai/v1",
            apiKey: "${MAXAPI_API_KEY}",
            api: "openai-completions",
            models: [
              {
                id: "YOUR_MODEL_ID",
                name: "MaxAPI model",
                input: ["text"],
                contextWindow: 128000,
                maxTokens: 8192
              }
            ]
          }
        }
      },
      agents: {
        defaults: {
          model: {
            primary: "maxapi/YOUR_MODEL_ID"
          }
        }
      }
    }
    ```

    <Note>
      Match `contextWindow` and `maxTokens` to the selected model. For a vision-capable model, set `input` to `["text", "image"]`.
    </Note>
  </Step>

  <Step title="Verify the configuration">
    Check model status, then select the MaxAPI model.

    ```bash theme={null}
    openclaw models status
    openclaw models list
    openclaw models set maxapi/YOUR_MODEL_ID
    ```
  </Step>
</Steps>

## Troubleshooting

<AccordionGroup>
  <Accordion title="Invalid API key">
    Confirm that `MAXAPI_API_KEY` is available to the shell or service that starts OpenClaw, and that the key is active.
  </Accordion>

  <Accordion title="Model not found">
    Ensure `models[].id` exactly matches the MaxAPI model list. Select it using the `maxapi/MODEL_ID` format.
  </Accordion>

  <Accordion title="The request returns 400">
    Ensure `api` is set to `openai-completions`. Check whether the model supports tools, image input, or other advanced request fields.
  </Accordion>

  <Accordion title="The request times out">
    Add or increase `timeoutSeconds` in the provider configuration. Use asynchronous endpoints for video, image, and music generation.
  </Accordion>
</AccordionGroup>

## Related documentation

* [OpenClaw model providers](https://docs.openclaw.ai/concepts/model-providers)
* [MaxAPI authentication](/en/guides/authentication)
* [MaxAPI API Reference](/en/api-reference/overview)