Serverless API is not available now for this model on any of the azure regions, please check again later.

Question

I cannot access my Mistral 2411 serverless api deployment anymore.

It worked until at least 2 days ago.

When I try to make a call with the endpoint and bearer token, I get:

Error: 403 RBAC: access denied Please check this guide to understand why this error code might have been returned

https://docs.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-online-endpoints#http-status-codes

I've never had this issue before, it's been deployed for a few months now.
Also through the playground I get the same error.

I tried deleting and re-creating the resource, but didn't help.

When trying to create the resource in a new project on the same hub, I see the following message:

'Serverless API is not available now for this model on any of the azure regions, please check again later.' But I can still subscribe and deploy without issues.

Is there a problem on Azure's side, or is there really a permission problem ? It seems weird to get this kind of error message if it's a server problem.
I've checked with the company Azure Admin, but no new permission policies have been added last few months.

Answer 1

Hi SIMONART Matthieu,

Thank you for posting your question in the Microsoft Q&A forum. 

It appears that your issue is likely due to a service disruption or policy change on Azure's side, rather than a local permission problem. The 403 RBAC error combined with the inability to deploy Mistral-24-11 in new projects suggests a platform-wide limitation. Since your admin confirms no recent policy changes, and the deployment option is now restricted, this points to Azure disabling Mistral-24-11 serverless APIs temporarily (possibly due to licensing, capacity, or compliance reasons). 

Please verify the following configurations to help resolve the access issues: 

1. Content Policy Compliance – Azure OpenAI models may temporarily restrict access if policy violations are detected. If this appears to be an error, please submit a support ticket to Azure and include the apim-request-id from the response headers for further investigation. 

Subscription & Region Availability – Certain Azure AI models require explicit permissions or may be region-restricted. Please confirm that your subscription has the necessary access and that the deployment region supports the model in use. 

Quota Allocation – Access restrictions can occur if usage exceeds allocated quotas. You may review and adjust quotas via the Azure portal as needed. 

API Key & Endpoint Accuracy – Some users have encountered issues due to incorrect URIs or keys. We recommend double-checking the API endpoint and authentication details for accuracy. 

If the issue persists after these checks, please review Azure Service Health for any ongoing incidents or contact Azure Support for further assistance. 

Let me know if you require any additional details, happy to help.  :)

 

If the above answer helped, please do not forget to "Accept Answer" as this may help other community members to refer the info if facing a similar issue. Your contribution to the Microsoft Q&A community is highly appreciated.