capacity

Query Azure OpenAI model capacity across regions and projects to rank deployment locations.

Updated Apr 12, 2026
One-click install
npx skills add https://github.com/davidrrowley/CortexYouV3 --skill capacity-davidrrowley
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: capacity
Source: https://github.com/davidrrowley/CortexYouV3/tree/main/.agents/skills/microsoft-foundry/models/deploy-model/capacity
Command: npx skills add https://github.com/davidrrowley/CortexYouV3 --skill capacity-davidrrowley

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires az, jq, python3, powershell, and includes scripts (resource) components.

What problem does it solve?

Finds where an Azure OpenAI model can actually be deployed by discovering platform capacity and subscription quota across regions and projects, removing guesswork when a target TPM or model availability is required.

Core Features & Use Cases

  • Capacity Discovery: Queries Azure modelCapacities to identify regions and SKUs with available capacity for a specified model and version.
  • Quota Annotation & Validation: Cross-references subscription quota per region and marks candidate locations with quota availability or limits.
  • Ranking & Project Suggestions: Ranks regions by target match, quota, project count, and available capacity and suggests sample projects when present.
  • Use Case: When you receive a quota error or need a region with X K TPM for gpt-4o, run discovery to get a ranked list of deployable regions and candidate projects before handing off to deployment flows.

Quick Start

Run the discover_and_rank script with your model name, version, and minimum TPM target to produce a ranked table of candidate regions and projects annotated with quota availability.

Frequently Asked Questions about capacity

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find Azure OpenAI capacity across regions when I hit a quota error?

To find Azure OpenAI capacity across regions after a quota error, you can query Azure modelCapacities and subscription usage. This cross-references available SKUs and quota limits to output a ranked table of deployable regions.

What is the best way to discover Azure regions for deploying specific OpenAI models?

The best way to discover Azure regions for deploying OpenAI models is to scan all accessible subscriptions and regions. This identifies model availability and ranks candidate deployment locations based on your specified capacity targets.

How do I check Azure OpenAI subscription quota and model availability before deployment?

To check Azure OpenAI subscription quota and model availability, you query the platform management API. This cross-references existing AI projects and subscription usage to validate candidate locations before deployment.

Do I need Azure CLI authentication to scan for OpenAI model capacity?

Yes, you need Azure CLI authentication and platform management API access to scan for OpenAI model capacity. These permissions are required to query modelCapacities and subscription usage across all accessible regions.

How are candidate regions ranked when searching for Azure OpenAI deployment capacity?

Candidate regions are ranked by target match, quota availability, project count, and available capacity. The resulting ranked table is annotated with quota information and suggests sample projects when present.

Can I use PowerShell and Python to automate Azure OpenAI region discovery?

Yes, you can use PowerShell and Python3 to automate Azure OpenAI region discovery. The skill utilizes these dependencies alongside Azure CLI and jq to execute discovery scripts and process capacity data.