Unable to deploy claude sonnet 5 model in microsoft foundry

Rohit Chamle 0 Reputation points
2026-10-13T05:38:38.4566667+00:00

Hello People,

I recently requested quota increase for Claude Sonnet 5 (from zero to 80000 TPM) and it got approved quickly. Although when I try to re-deploy the model in Microsoft foundry it is not being provisioned. Am I missing something here?

Any help or assistant is appreciated.

Thanks, Rohit

Microsoft Foundry
Microsoft Foundry

A unified Azure platform for creating and managing AI models, agents, and applications with built‑in enterprise security, monitoring, and governance

0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest

  1. AI answer

    2026-10-13T05:39:02.3066667+00:00

    If quota was approved but the Claude Sonnet 5 deployment still is not provisioning, the most likely causes are region availability, quota propagation delay, or a backend capacity/provisioning issue rather than a missing deployment setting.

    Try this end-to-end check:

    1. In Microsoft Foundry, go to Manage > Quota and confirm the approved quota is visible for the correct subscription.
      • Quota changes can take up to 15 minutes to propagate after approval.
      • Make sure the deployment is being created in the same subscription where the quota was approved.
    2. Verify the target region supports the Claude family.
      • If the selected region is not supported, deployment can fail.
      • Claude model availability is region-dependent.
    3. If quota appears full even though there are no active deployments, check for deleted resources still holding quota.
      • Soft-deleted Foundry or AI Services accounts can reserve TPM quota for up to 48 hours.
      • If applicable, list deleted accounts with: 
             az cognitiveservices account list-deleted -o table
      • Purge accounts that are no longer needed.
      • Warning: purging deleted accounts is destructive and permanently removes those soft-deleted resources.
    4. Retry with a lower capacity allocation first.
      • If the deployment was attempted at the full approved TPM, try assigning a smaller amount initially and then scale up.
      • This is a documented troubleshooting step when quota is insufficient or appears blocked.
    5. If the deployment status shows Failed, recheck both conditions together:
      • the model is available in the selected region
      • sufficient quota is actually available to that deployment
    6. If the deployment still does not provision after quota is visible and the region is supported, treat it as a backend provisioning/capacity issue.
      • Claude deployments can fail because capacity cannot be allocated for that subscription/region combination.
      • In that case, the practical next step is to contact Azure support.

    A few important points for this scenario:

    • Approved quota alone does not guarantee immediate successful provisioning.
    • Deleting unrelated model deployments does not necessarily help unless they are consuming the same Claude quota allocation.
    • If the portal shows the quota increase but provisioning still fails, that strongly points to service-side capacity or entitlement checks.
    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.