What problem does it solve? Finding an Azure region with enough available capacity and subscription quota to deploy an Azure OpenAI model is a manual, error-prone process involving multiple REST API calls and quota lookups. This Skill automates that discovery and ranks candidate regions so you can pick a deployment target with confidence. ## Core Features & Use Cases - Cross-Region Capacity Discovery: Queries the Azure Cognitive Services modelCapacities API across all regions for a given model and version, filtering by SKU such as GlobalStandard. - Quota Validation: Checks subscription-level quota per region via az cognitiveservices usage list and annotates results so regions at their limit are flagged. - Ranked Recommendations: Cross-references existing AI Foundry projects and outputs a ranked table sorted by capacity target, quota availability, and project count. - Use Case: You need to deploy gpt-4o with at least 100K TPM. Run the discovery script to get a ranked table of regions showing available capacity, remaining quota, and existing projects, then hand off to a deployment skill. ## Quick Start Ask the agent to find which Azure regions have enough capacity and quota to deploy your chosen model, for example by saying find the best region to deploy gpt-4o with at least 100K TPM.
