Edit

Rate limits, region support, and enterprise features for evaluation (classic)

Currently viewing: Foundry (classic) portal version - Switch to version for the new Foundry portal

This article provides an overview of which regions support AI-assisted evaluators, the rate limits that apply to evaluation runs, how to configure virtual network support for network isolation, and using your own storage account to run evaluations.

Regional availability

Supported regions for Agent playground evaluations

The agent playground evaluations are supported in the following regions:

Americas Europe
East US 2 France Central
West US Norway East
West US 2 Sweden Central
West US 3 Germany West Central
Central US Italy North
East US Poland Central
North Central US Spain Central
South Central US

Supported regions for batch evaluations

The batch evaluations are supported in the following regions:

Americas Europe Asia Pacific Middle East & Africa
Brazil South France Central Australia East South Africa North
Canada Central Germany West Central Central India UAE North
Canada East Italy North East Asia
Central US North Europe Japan East
East US Norway East Japan West
East US 2 Poland Central Korea Central
North Central US Spain Central South India
South Central US Sweden Central Southeast Asia
West Central US Switzerland North
West US UK South
West US 2 West Europe
West US 3

Risk and safety evaluators and AI red teaming region support

The following safety evaluators and AI red teaming are supported in these regions: Hate and unfairness, Sexual, Violent, Self-harm, Indirect attack, Code vulnerabilities, Ungrounded attributes, and AI red teaming.

Americas Europe
East US 2 France Central
North Central US Sweden Central
Switzerland West

Supported regions for Groundedness Pro:

  • East US 2
  • Sweden Central

Supported regions for Protected material:

  • East US 2

Azure OpenAI graders regional availability

For the Azure OpenAI graders regional list, see Regional availability.

Rate limits

The following rate limits apply to evaluation runs:

Limit Value
Maximum size per row 2 MB
Maximum rows per batch evaluation 100,000

Evaluation run creations are rate-limited at the tenant, subscription, and project levels. If you exceed the limit:

  • The response includes a retry-after header with the wait time.
  • The response body contains rate limit details.

Use exponential backoff when retrying failed requests.

Virtual network support for evaluation

For network isolation, you can bring your own virtual network for evaluation. To learn more, see How to configure a private link.

Virtual network support for evaluation requires network injection (subnet delegation), but if you only need evaluation capabilities and do not require full agent support (Cosmos DB, AI Search, or project capability host), consider using the simplified evaluation-only setup template (15a) instead. It deploys a minimal network-secured environment tailored for evaluation scenarios with fewer resources and reduced complexity.

Note

If you connect Application Insights, evaluation data is sent to it.

Important

To prevent evaluation and red teaming run failures, assign the Foundry User role to the project's Managed Identity during initial project setup.

Important

The Foundry RBAC roles were recently renamed. Foundry User, Foundry Owner, Foundry Account Owner, and Foundry Project Manager were previously named Azure AI User, Azure AI Owner, Azure AI Account Owner, and Azure AI Project Manager. You might still see the previous names in some places while the rename rolls out. The role IDs and core permissions are unchanged by the rename.

Virtual network region support

Bringing your own virtual network for evaluation is supported in all regions except:

  • Central India
  • East Asia
  • North Europe
  • Qatar Central

Bring your own storage

You can use your own storage account to run evaluations.

You can either use a Bicep template or manually create and provision access to your storage account in the Azure portal. To use a Bicep template, follow these steps.

  1. Create and connect your storage account to your Foundry project at the resource level. You can use a Bicep template, which provisions and connects a storage account to your Foundry project with key authentication.
  2. Make sure the connected storage account has access to all projects.
  3. If you connected your storage account by using Microsoft Entra ID, make sure to give managed identity Storage Blob Data Owner permissions to both your account and the Foundry project resource in the Azure portal.