You can change the AI model used by your OpenClaw by updating your API provider settings in hPanel. With Managed OpenClaw, you can use nexos.ai, OpenAI, Anthropic, or Gemini as your provider, and pick different models within each.
How to change the AI model
To use a model from a different provider (possibilities are OpenAI, Anthropic, or xAI), you can manually update your API key.
- Go to the OpenClaw section in hPanel.
- Click Manage next to your application.
- In the left sidebar, click AI provider.
- Select the provider tab for your chosen vendor (nexos.ai, OpenAI, Anthropic, or Gemini).
- Enter the API key obtained from your chosen vendor (e.g., OpenAI, Anthropic, Google Gemini, or nexos.ai).
NOTE: Without an API key and purchased credits, OpenClaw will not be able to function properly. - Once the API key is saved, the AI Model dropdown becomes available. Select the model you want to use.
- Save the changes.
After adding a new API key, click OpenClaw to launch the application. A Starting OpenClaw window will appear while the system automatically identifies the new key and runs the necessary configurations.
Once the process is complete, your assistant will be ready for use with the new model.
Connect via OAuth (OpenAI)
If you have a ChatGPT Plus or Pro subscription, you can connect it directly instead of using an API key. Credits are drawn from your connected ChatGPT subscription, and tokens are refreshed automatically.
- Go to the OpenClaw section in hPanel.
- Click Manage next to your application.
- In the left sidebar, click AI provider, then select the OpenAI tab.
- Under Use your ChatGPT subscription, click Connect.
- A window will appear with a link to https://auth.openai.com/codex/device and a verification code. Open the link, sign in with your ChatGPT Plus or Pro account, enter the code shown in hPanel, and complete verification.
- Once connected, your application will restart automatically to apply the connection.
