1---
2name: capacity
3description: "Discovers available Azure OpenAI model capacity across regions and projects. Analyzes quota limits, compares availability, and recommends optimal deployment locations based on capacity requirements. USE FOR: find capacity, check quota, where can I deploy, capacity discovery, best region for capacity, multi-project capacity search, quota analysis, model availability, region comparison, check TPM availability. DO NOT USE FOR: actual deployment (hand off to preset or customize after discovery), quota increase requests (direct user to Azure Portal), listing existing deployments."
4license: MIT
5metadata:
6  author: Microsoft
7  version: "1.0.0"
8---
9 
10# Capacity Discovery
11 
12Finds available Azure OpenAI model capacity across all accessible regions and projects. Recommends the best deployment location based on capacity requirements.
13 
14## Quick Reference
15 
16| Property | Description |
17|----------|-------------|
18| **Purpose** | Find where you can deploy a model with sufficient capacity |
19| **Scope** | All regions and projects the user has access to |
20| **Output** | Ranked table of regions/projects with available capacity |
21| **Action** | Read-only analysis — does NOT deploy. Hands off to preset or customize |
22| **Authentication** | Azure CLI (`az login`) |
23 
24## When to Use This Skill
25 
26- ✅ User asks "where can I deploy gpt-4o?"
27- ✅ User specifies a capacity target: "find a region with 10K TPM for gpt-4o"
28- ✅ User wants to compare availability: "which regions have gpt-4o available?"
29- ✅ User got a quota error and needs to find an alternative location
30- ✅ User asks "best region and project for deploying model X"
31 
32**After discovery → hand off to [preset](../preset/SKILL.md) or [customize](../customize/SKILL.md) for actual deployment.**
33 
34## Scripts
35 
36Pre-built scripts handle the complex REST API calls and data processing. Use these instead of constructing commands manually.
37 
38| Script | Purpose | Usage |
39|--------|---------|-------|
40| `scripts/discover_and_rank.ps1` | Full discovery: capacity + projects + ranking | Primary script for capacity discovery |
41| `scripts/discover_and_rank.sh` | Same as above (bash) | Primary script for capacity discovery |
42| `scripts/query_capacity.ps1` | Raw capacity query (no project matching) | Quick capacity check or version listing |
43| `scripts/query_capacity.sh` | Same as above (bash) | Quick capacity check or version listing |
44 
45## Workflow
46 
47### Phase 1: Validate Prerequisites
48 
49```bash
50az account show --query "{Subscription:name, SubscriptionId:id}" --output table
51```
52 
53### Phase 2: Identify Model and Version
54 
55Extract model name from user prompt. If version is unknown, query available versions:
56 
57```powershell
58.\scripts\query_capacity.ps1 -ModelName <model-name>
59```
60```bash
61./scripts/query_capacity.sh <model-name>
62```
63 
64This lists available versions. Use the latest version unless user specifies otherwise.
65 
66### Phase 3: Run Discovery
67 
68Run the full discovery script with model name, version, and minimum capacity target:
69 
70```powershell
71.\scripts\discover_and_rank.ps1 -ModelName <model-name> -ModelVersion <version> -MinCapacity <target>
72```
73```bash
74./scripts/discover_and_rank.sh <model-name> <version> <min-capacity>
75```
76 
77> 💡 The script automatically queries capacity across ALL regions, cross-references with the user's existing projects, and outputs a ranked table sorted by: meets target → project count → available capacity.
78 
79### Phase 3.5: Validate Subscription Quota
80 
81After discovery identifies candidate regions, validate that the user's subscription actually has available quota in each region. Model capacity (from Phase 3) shows what the platform can support, but subscription quota limits what this specific user can deploy.
82 
83```powershell
84# For each candidate region from discovery results:
85$usageData = az cognitiveservices usage list --location <region> --subscription $SUBSCRIPTION_ID -o json 2>$null | ConvertFrom-Json
86 
87# Check quota for each SKU the model supports
88# Quota names follow pattern: OpenAI.<SKU>.<model-name>
89$usageEntry = $usageData | Where-Object { $_.name.value -eq "OpenAI.<SKU>.<model-name>" }
90 
91if ($usageEntry) {
92  $quotaAvailable = $usageEntry.limit - $usageEntry.currentValue
93} else {
94  $quotaAvailable = 0  # No quota allocated
95}
96```
97```bash
98# For each candidate region from discovery results:
99usage_json=$(az cognitiveservices usage list --location <region> --subscription "$SUBSCRIPTION_ID" -o json 2>/dev/null)
100 
101# Extract quota for specific SKU+model
102quota_available=$(echo "$usage_json" | jq -r --arg name "OpenAI.<SKU>.<model-name>" \
103  '.[] | select(.name.value == $name) | .limit - .currentValue')
104```
105 
106**Annotate discovery results:**
107 
108Add a "Quota Available" column to the ranked output from Phase 3:
109 
110| Region | Available Capacity | Meets Target | Projects | Quota Available |
111|--------|-------------------|--------------|----------|-----------------|
112| eastus2 | 120K TPM | ✅ | 3 | ✅ 80K |
113| westus3 | 90K TPM | ✅ | 1 | ❌ 0 (at limit) |
114| swedencentral | 100K TPM | ✅ | 0 | ✅ 100K |
115 
116Regions/SKUs where `quotaAvailable = 0` should be marked with ❌ in the results. If no region has available quota, hand off to the [quota skill](../../../quota/quota.md) for increase requests and troubleshooting.
117 
118### Phase 4: Present Results and Hand Off
119 
120After the script outputs the ranked table (now annotated with quota info), present it to the user and ask:
121 
1221. 🚀 **Quick deploy** to top recommendation with defaults → route to [preset](../preset/SKILL.md)
1232. ⚙️ **Custom deploy** with version/SKU/capacity/RAI selection → route to [customize](../customize/SKILL.md)
1243. 📊 **Check another model** or capacity target → re-run Phase 2
1254. ❌ Cancel
126 
127### Phase 5: Confirm Project Before Deploying
128 
129Before handing off to preset or customize, **always confirm the target project** with the user. See the [Project Selection](../SKILL.md#project-selection-all-modes) rules in the parent router.
130 
131If the discovery table shows a sample project for the chosen region, suggest it as the default. Otherwise, query projects in that region and let the user pick.
132 
133## Error Handling
134 
135| Error | Cause | Resolution |
136|-------|-------|------------|
137| "No capacity found" | Model not available or all at quota | Hand off to [quota skill](../../../quota/quota.md) for increase requests and troubleshooting |
138| Script auth error | `az login` expired | Re-run `az login` |
139| Empty version list | Model not in region catalog | Try a different region: `./scripts/query_capacity.sh <model> "" eastus` |
140| "No projects found" | No AI Services resources | Guide to `project/create` skill or Azure Portal |
141 
142## Related Skills
143 
144- **[preset](../preset/SKILL.md)** — Quick deployment after capacity discovery
145- **[customize](../customize/SKILL.md)** — Custom deployment after capacity discovery
146- **[quota](../../../quota/quota.md)** — For quota viewing, increase requests, and troubleshooting quota errors, defer to this skill instead of duplicating guidance
147