openclaw - 💡(How to fix) Fix [Bug]: Control UI Model Selection Returns "model not allowed" Error - Cannot Switch Custom Provider Models [3 comments, 3 participants]

When switching to a model configured under a custom Provider (e.g., nvidia, modelscope) in the Control UI, the "model not allowed" error is returned. 4. Observe the error message Error returned: Failed to set model: GatewayRequestError: model not allowed: z-ai/glm5

TL;DR

The issue can be resolved by ensuring that the model keys in agents.defaults.models are correctly prefixed with the provider name and configured in openclaw.json.

Guidance

• Verify that the openclaw.json file contains the correct configuration for the custom Provider (e.g., nvidia) and that the model keys in agents.defaults.models are prefixed with the provider name.
• Check the model selector in the Control UI to ensure that the model is correctly listed and selectable.
• Confirm that the GatewayRequestError is not related to a networking or permissions issue.
• Review the openclaw.json file for any typos or incorrect configurations that may be causing the "model not allowed" error.

Example

No code snippet is provided as the issue does not contain sufficient information to create a specific example.

Notes

The issue may be related to a configuration error in openclaw.json or agents.defaults.models, and resolving this issue may require careful review of the configuration files.

Recommendation

Apply workaround: Verify and correct the configuration files (openclaw.json and agents.defaults.models) to ensure that the model keys are correctly prefixed with the provider name and that the custom Provider is correctly configured.

FAIL-SAFE

If the issue persists after verifying the configuration files, further investigation into the GatewayRequestError and the custom Provider configuration may be necessary.