1# Detailed Workflows: Quota Management
2 
3**Table of Contents:** [Workflow 1: View Current Quota Usage](#workflow-1-view-current-quota-usage---detailed-steps) · [Workflow 2: Find Best Region for Model Deployment](#workflow-2-find-best-region-for-model-deployment---detailed-steps) · [Workflow 3: Check Quota Before Deployment](#workflow-3-check-quota-before-deployment---detailed-steps) · [Workflow 4: Monitor Quota Across Deployments](#workflow-4-monitor-quota-across-deployments---detailed-steps) · [Quick Command Reference](#quick-command-reference) · [MCP Tools Reference](#mcp-tools-reference-optional-wrappers)
4 
5## Workflow 1: View Current Quota Usage - Detailed Steps
6 
7### Step 1: Show Regional Quota Summary (REQUIRED APPROACH)
8 
9> **CRITICAL AGENT INSTRUCTION:**
10> - When showing quota: Query REGIONAL quota summary, NOT individual resources
11> - DO NOT run `az cognitiveservices account list` for quota queries
12> - DO NOT filter resources by username or name patterns
13> - ONLY check specific resource deployments if user provides resource name
14> - Quotas are managed at SUBSCRIPTION + REGION level, NOT per-resource
15 
16**Show Regional Quota Summary:**
17 
18```bash
19# Get subscription ID
20subId=$(az account show --query id -o tsv)
21 
22# Check quota for key regions
23regions=("eastus" "eastus2" "westus" "westus2")
24for region in "${regions[@]}"; do
25  echo "=== Region: $region ==="
26  az rest --method get \
27    --url "https://management.azure.com/subscriptions/$subId/providers/Microsoft.CognitiveServices/locations/$region/usages?api-version=2023-05-01" \
28    --query "value[?contains(name.value,'OpenAI.Standard')].{Model:name.value, Used:currentValue, Limit:limit, Available:(limit-currentValue)}" \
29    --output table
30  echo ""
31done
32```
33 
34### Step 2: If User Asks for Specific Resource (ONLY IF EXPLICITLY REQUESTED)
35 
36```bash
37# User must provide resource name
38az cognitiveservices account deployment list \
39  --name <user-provided-resource-name> \
40  --resource-group <user-provided-rg> \
41  --query '[].{Name:name, Model:properties.model.name, Capacity:sku.capacity, SKU:sku.name}' \
42  --output table
43```
44 
45**Alternative - Use MCP Tools (Optional Wrappers):**
46```
47foundry_models_deployments_list(
48  resource-group="<rg>",
49  azure-ai-services="<resource-name>"
50)
51```
52*Note: MCP tools are convenience wrappers around the same control plane APIs shown above.*
53 
54**Interpreting Results:**
55- `Used` (currentValue): Currently allocated quota
56- `Limit`: Maximum quota available in region
57- `Available`: Calculated as `limit - currentValue`
58 
59## Workflow 2: Find Best Region for Model Deployment - Detailed Steps
60 
61### Step 1: Check Single Region
62 
63```bash
64# Get subscription ID
65subId=$(az account show --query id -o tsv)
66 
67# Check quota for GPT-4o Standard in a specific region
68region="eastus"  # Change to your target region
69az rest --method get \
70  --url "https://management.azure.com/subscriptions/$subId/providers/Microsoft.CognitiveServices/locations/$region/usages?api-version=2023-05-01" \
71  --query "value[?name.value=='OpenAI.Standard.gpt-4o'].{Model:name.value, Used:currentValue, Limit:limit, Available:(limit-currentValue)}" \
72  -o table
73```
74 
75### Step 2: Check Multiple Regions (Common Regions)
76 
77Check these regions in sequence by changing the `region` variable:
78- `eastus`, `eastus2` - US East Coast
79- `westus`, `westus2`, `westus3` - US West Coast
80- `swedencentral` - Europe (Sweden)
81- `canadacentral` - Canada
82- `uksouth` - UK
83- `japaneast` - Asia Pacific
84 
85**Alternative - Use MCP Tool:**
86```
87model_quota_list(region="eastus")
88```
89Repeat for each target region.
90 
91**Key Points:**
92- Query returns `currentValue` (used), `limit` (max), and calculated `Available`
93- Standard SKU format: `OpenAI.Standard.<model-name>`
94- For PTU: `OpenAI.ProvisionedManaged.<model-name>`
95- Focus on 2-3 regions relevant to your location rather than checking all regions
96 
97## Workflow 3: Check Quota Before Deployment - Detailed Steps
98 
99**Steps:**
1001. Check current usage (workflow #1)
1012. Calculate available: `limit - currentValue`
1023. Compare: `available >= required_capacity`
1034. If insufficient: Use workflow #2 to find region with capacity, or request increase
104 
105## Workflow 4: Monitor Quota Across Deployments - Detailed Steps
106 
107**Recommended Approach - Regional Quota Overview:**
108 
109Show quota by region (better than listing all resources):
110 
111```bash
112subId=$(az account show --query id -o tsv)
113regions=("eastus" "eastus2" "westus" "westus2" "swedencentral")
114 
115for region in "${regions[@]}"; do
116  echo "=== Region: $region ==="
117  az rest --method get \
118    --url "https://management.azure.com/subscriptions/$subId/providers/Microsoft.CognitiveServices/locations/$region/usages?api-version=2023-05-01" \
119    --query "value[?contains(name.value,'OpenAI')].{Model:name.value, Used:currentValue, Limit:limit, Available:(limit-currentValue)}" \
120    --output table
121  echo ""
122done
123```
124 
125**Alternative - Check Specific Resource:**
126 
127If user wants to monitor a specific resource, ask for resource name first:
128 
129```bash
130# List deployments for specific resource
131az cognitiveservices account deployment list \
132  --name <resource-name> \
133  --resource-group <rg> \
134  --query '[].{Name:name, Model:properties.model.name, Capacity:sku.capacity}' \
135  --output table
136```
137 
138> **Note:** Don't automatically iterate through all resources in the subscription. Show regional quota summary or ask for specific resource name.
139 
140## Quick Command Reference
141 
142```bash
143# View quota for specific model using REST API
144subId=$(az account show --query id -o tsv)
145region="eastus"  # Change to your region
146az rest --method get \
147  --url "https://management.azure.com/subscriptions/$subId/providers/Microsoft.CognitiveServices/locations/$region/usages?api-version=2023-05-01" \
148  --query "value[?contains(name.value,'gpt-4')].{Name:name.value, Used:currentValue, Limit:limit, Available:(limit-currentValue)}" \
149  --output table
150 
151# List all deployments with capacity
152az cognitiveservices account deployment list \
153  --name <resource-name> \
154  --resource-group <rg> \
155  --query '[].{Name:name, Model:properties.model.name, Capacity:sku.capacity}' \
156  --output table
157 
158# Delete deployment to free quota
159az cognitiveservices account deployment delete \
160  --name <resource-name> \
161  --resource-group <rg> \
162  --deployment-name <deployment-name>
163```
164 
165## MCP Tools Reference (Optional Wrappers)
166 
167**Note:** All quota operations are control plane (management) operations. MCP tools are optional convenience wrappers around Azure CLI commands.
168 
169| Tool | Purpose | Equivalent Azure CLI |
170|------|---------|---------------------|
171| `foundry_models_deployments_list` | List all deployments with capacity | `az cognitiveservices account deployment list` |
172| `model_quota_list` | List quota and usage across regions | `az rest` (Management API) |
173| `model_catalog_list` | List available models from catalog | `az rest` (Management API) |
174| `foundry_resource_get` | Get resource details and endpoint | `az cognitiveservices account show` |
175 
176**Recommended:** Use Azure CLI commands directly for control plane operations.
177