# Underutilized PTU quota for Azure OpenAI deployments Canonical: https://www.pointfive.co/efficiency-hub/inefficiencies/underutilized-ptu-quota-for-azure-openai-deployments-da831 When organizations size PTU capacity based on peak expectations or early traffic projections, they often end up with more throughput than regularly... By: Ariel Fishman-Lichterman Updated: 2026-09-28 [Cloud Efficiency Hub](https://www.pointfive.co/efficiency-hub) # Underutilized PTU Quota for Azure OpenAI Deployments The short version When organizations size PTU capacity based on peak expectations or early traffic projections, they often end up with more throughput than regularly required. [Ariel Fishman-Lichterman](https://www.linkedin.com/in/ariel-fishman-lichterman-01a096214/) Cloud Researcher @ PointFive Azure service [Azure Cognitive Services](https://www.pointfive.co/efficiency-hub/cloud-services/azure-cognitive-services) Category [AI](https://www.pointfive.co/efficiency-hub/service-category/ai) Reference CER-0266 Type Overprovisioned Resource ## Explanation Why the waste happens and who it affects. If real-world usage plateaus below provisioned levels, a portion of the PTU capacity remains idle but still generates full spend each hour. This is especially common shortly after production launch or during adoption of newly released model versions, where early conservative sizing leads to long-term over-allocation. Rightsizing PTUs based on observed usage patterns ensures that capacity matches actual demand. ## Billing model The pricing dimensions that drive this cost. PTU pricing is based on the number of provisioned throughput units, not actual usage. Underutilized PTUs still incur full hourly charges, making over-allocation a direct source of avoidable cost. ## How to detect 4 checks to find it in your estate. - Review PTU deployments for consistently low or flat throughput utilization over representative time periods - Compare provisioned PTU levels against actual workload demand to identify idle capacity - Identify deployments sized for initial peak estimates that no longer match steady-state usage - Evaluate whether recent model or workload changes have altered throughput requirements ## How to fix 4 ways to remove the waste. - Reduce PTU allocations to align with actual utilization while preserving required performance levels. If the PTUs are covered by a reservation, pair the scale-down with a reservation exchange or cancellation (fees and limits apply), otherwise the reservation keeps billing its original quantity and there are no savings; released capacity may also not be available again later - Implement recurring rightsizing reviews to adjust PTU levels as workload patterns evolve - Use workload performance testing to validate that reduced capacity meets latency and throughput goals - Consider shifting variable or declining workloads from PTUs to PAYG where appropriate ## Documentation Vendor references for pricing and configuration. - [Provisioned throughput billing and cost management learn.microsoft.com](https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/provisioned-throughput-billing) - [Save costs with Microsoft Foundry Provisioned Throughput Reservations learn.microsoft.com](https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/microsoft-foundry) ## Related inefficiencies [Browse the library](https://www.pointfive.co/efficiency-hub) - Azure Cognitive Services CER-0238 ### [Always-On PTUs for Seasonal or Cyclical Azure OpenAI Workloads](https://www.pointfive.co/efficiency-hub/inefficiencies/always-on-ptus-for-seasonal-or-cyclical-azure-openai-workloads-ccb97) Many Azure OpenAI workloads - such as reporting pipelines, marketing workflows, batch inference jobs, or time-bound customer interactions-only run during specific periods. When PTUs remain fully provisioned 24/7, organizations incur... AI - Azure Cognitive Services CER-0248 ### [Non-Production Azure OpenAI Deployments Using PTUs Instead of PAYG](https://www.pointfive.co/efficiency-hub/inefficiencies/non-production-azure-openai-deployments-using-ptus-instead-of-payg-9fe81) Development, testing, QA, and sandbox environments rarely have the steady, predictable traffic patterns needed to justify PTU deployments. These workloads often run intermittently, with lower throughput and shorter usage windows. When PTUs... AI - Azure Cognitive Services CER-0246 ### [Missing Reserved PTUs for Steady-State Azure OpenAI Workloads](https://www.pointfive.co/efficiency-hub/inefficiencies/missing-reserved-ptus-for-steady-state-azure-openai-workloads-f78d4) Many production Azure OpenAI workloads - such as chatbots, inference services, and retrieval-augmented generation (RAG) pipelines-use PTUs consistently throughout the day. When usage stabilizes after initial experimentation, continuing to... AI --- Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.