Azure AI Foundry Model Deployments Lister

This Python script lists all Azure AI Foundry model deployments across all your Azure subscriptions and outputs the results as both a formatted table and CSV file for reporting.

The Microsoft Foundry control plane UI currently only allows viewing deployments by clicking through subscriptions one at a time. This script provides a programmatic way to view all your model deployments across all subscriptions in a single report.

• Scans all accessible Azure subscriptions
• Lists all AI Foundry model deployments
• Outputs formatted console table
• Exports to CSV for reporting
• Matches the UI column structure shown in the Foundry portal

1. Python 3.7 or higher
2. Azure CLI (for authentication) - Install from: https://learn.microsoft.com/cli/azure/install-azure-cli
3. Azure credentials with appropriate permissions:
   • Reader access to subscriptions
   • Access to AI Foundry resources

1. Clone or download this repository

2. Install the required Python packages:

   pip install -r requirements.txt

3. Authenticate with Azure CLI:

   az login

Simply run the script:

python list_foundry_deployments.py

The script will:

1. Authenticate using your Azure credentials
2. Discover all accessible subscriptions
3. Scan each subscription for AI Foundry resources
4. Query each resource for model deployments
5. Display results in a formatted table
6. Export to a timestamped CSV file (e.g., foundry_deployments_20250425_143022.csv)

The script displays a formatted table with key columns:

• Name
• Project
• Version
• State
• Base model
• PTU capacity
• Rate limit (tokens per minute)
• Subscription

The CSV file includes all columns matching the Foundry UI:

• Name
• Project
• Version
• State
• Guardrails
• Deployment type
• PTU capacity
• Rate limit (tokens per minute)
• Base model
• Retirement date
• Fine-tuned
• Model cost
• Deployed
• Subscription
• Resource Group
• Model Publisher
• Connection Name

The script uses DefaultAzureCredential which tries multiple authentication methods in order:

1. Environment variables
2. Managed Identity
3. Azure CLI authentication
4. Azure PowerShell
5. Interactive browser

For most users, running az login before executing the script is sufficient.

• Ensure you have AI Foundry projects with active deployments
• Verify you have sufficient permissions (Reader role at minimum)
• Check that you're logged in with the correct Azure account

• Run az login to authenticate
• Run az account list to verify you can access subscriptions
• Ensure your account has access to the subscriptions containing Foundry resources

• The script uses the AI Foundry REST API (v1)
• Some resources may not expose the deployments endpoint
• The script will skip resources it cannot access and continue

• Only enabled subscriptions are scanned
• The script uses parallel processing (max 20 subscriptions concurrently, 10 accounts per subscription) for performance at scale
• All data comes directly from Azure APIs with no assumptions or fallbacks

No Hardcoded Fallbacks:

• The script uses only data returned by the Azure API
• Empty or missing fields are displayed as empty strings, not placeholder values
• This ensures accurate reporting of actual deployment states

Rate Limits:

• Extracted directly from the API's rateLimits field
• The API may return multiple rate limit entries or none at all
• The script displays the first valid count value found, or -- if unavailable
• Rate limit structure varies by deployment type and may not always be populated

Fine-tuned Detection:

• Identified by checking if the model name starts with ft: prefix
• This is a heuristic and may not catch all fine-tuned models if they follow different naming conventions
• Azure's fine-tuning naming format: ft:base-model:org:custom-suffix:id

PTU Capacity:

• Extracted directly from sku.capacity in the API response
• May be empty/null for non-PTU deployments (e.g., pay-as-you-go deployments)

Deployment Type:

• Uses the exact SKU name returned by the API (e.g., "Standard", "GlobalStandard", "ProvisionedManaged")
• No transformation or interpretation applied

Empty Fields:

• If the API doesn't provide a value, the field will be empty in both console and CSV output
• This applies to: State, Guardrails, Version Upgrade Option, and other optional fields

This script uses the Azure AI Foundry Deployments API:

• Endpoint: https://learn.microsoft.com/en-us/rest/api/aifoundry/aiprojects/deployments/list
• API Version: v1