PTU sizing and throughput increment in Gpt-4o-mini

Hi ,

Thanks for reaching out to Microsoft Q&A.

You cannot independently increase token throughput for gpt-4o-mini beyond what your current PTU allocation allows. Throughput is strictly bound to PTUs assigned to the model.

How throughput actually scales:

• Token per minute and requests per second scale linearly only when Microsoft assigns additional PTUs.
• Raising a quota request alone does not guarantee allocation. PTUs are capacity-constrained and region-dependent.

Microsoft approval for PTU increases can take weeks.

There is no escalation path other than Azure support.

If you need immediate high throughput, GPT-4o-mini may not meet burst requirements today.

If this is prod critical and latency or throughput is tight, design with fallback models or multi-region deployments. Do not assume PTU requests will be approved quickly. 

Please 'Upvote'(Thumbs-up) and 'Accept' as answer if the reply was helpful. This will be benefitting other community members who face the same issue.

Hi SysAdmin

Did you get any chance to review the above response.please find the details requested through private message

Thank you!