Share via

while testing an agent in AI Foundry an error response is returned for the prompt

Shariq Hashmi 10 Reputation points
2025-12-13T09:46:23.4733333+00:00

While testing an agent in playground in AI Foundry (New Foundry) agent starts responding to the prompt but before completing response it throws below error. It is 100% recurring on more than 5 attempts, sharing two errors with request id below.

Attempt 1: ErrorAn error occurred while processing your request. You can retry your request, or contact us through an Azure support request at: https://go.microsoft.com/fwlink/?linkid=2213926 if the error persists. Please include the request ID 68ed3d2a-ffb0-4d1e-b608-b8cbf058df32 in your message.

Attempt 2: ErrorAn error occurred while processing your request. You can retry your request, or contact us through an Azure support request at: https://go.microsoft.com/fwlink/?linkid=2213926 if the error persists. Please include the request ID 9e2f7fba-b4a5-45a5-8d6a-1efc077e31ac in your message.

Foundry Tools
Foundry Tools

Formerly known as Azure AI Services or Azure Cognitive Services is a unified collection of prebuilt AI capabilities within the Microsoft Foundry platform

0 comments

3 answers

Sort by: Most helpful
Most helpful Newest Oldest
  1. Shariq Hashmi 10 Reputation points
    2025-12-14T21:27:58.82+00:00

    can Microsoft support team please look into this issue? I am continuously facing this issue, Ai answer is not helpful

    Was this answer helpful?

    1 person found this answer helpful.
  2. SRILAKSHMI C 19,090 Reputation points Microsoft External Staff Moderator
    2025-12-23T14:26:59.64+00:00

    Hi Shariq Hashmi,

    Thank you for sharing the details,

    Based on the behavior you’re seeing where the agent starts responding and then consistently fails mid-generation this does not appear to be a prompt issue. This pattern typically points to a service-side or configuration-related issue in the AI Foundry Agent Playground.

    Below are the most relevant checks and explanations to help narrow this down.

    What the error indicates

    The generic error returned with a request ID usually means the request reached the backend, but an internal failure occurred during response generation. Since this is 100% reproducible, retries alone will not resolve it.

    This can be caused by either:

    • A backend service issue, or
    • A misconfiguration in the AI Foundry project or agent setup

    Things to verify on your side

    1. Permissions and role assignments

    Ensure the identity used by the agent (user, service principal, or managed identity) has the correct roles on the AI Foundry project:

    Azure AI User (minimum)

    Azure AI Project Manager (recommended)

    You can verify this under Access Control (IAM) for the AI project.

    1. Networking configuration

    If your project uses:

    VNET integration

    Private endpoints

    Restricted network access

    Please confirm:

    All related resources (AI project, networking components) are in the same region

    Networking was configured at project creation time

    Adding or changing VNET / private endpoint settings after agent creation is not supported and can lead to runtime failures like the one you’re seeing.

    1. API key and connection settings

    Confirm that:

    The API key or managed identity used has access to the Azure OpenAI / AI Foundry resource

    Required connections (for example, model connections) exist and are active

    If agents were deployed via ARM/Bicep, ensure the template includes the required:

    aiServicesConnections configuration

    Missing this can cause agents to fail at runtime even though creation succeeds.

    1. Quotas and limits

    Although the error message is generic, quota exhaustion can sometimes surface this way.

    Please check:

    Agent limits per project/region

    Model quota availability

    Token or request limits

    You can verify this in the AI Foundry quota and limits blade.

    1. Model and region compatibility

    Ensure that:

    The model selected for the agent is supported in your region

    The model is currently available and healthy

    Using an unsupported or constrained model can result in mid-response failures.

    When this is likely a service-side issue

    Given that:

    The response begins streaming

    The failure is consistent

    Multiple attempts return different request IDs

    This strongly suggests a backend execution issue in the Foundry Agent Playground.

    Recommended steps

    Open an Azure Support Request Include:

    • Request IDs:
      • 68ed3d2a-ffb0-4d1e-b608-b8cbf058df32
      • 9e2f7fba-b4a5-45a5-8d6a-1efc077e31ac
    • Subscription ID
    • Tenant ID
    • AI Foundry project name and region
    • Confirmation that the error occurs mid-response in the Playground

    Test via API If the agent is callable via REST/SDK, test outside the Playground to determine whether the issue is UI-specific or service-wide.

    Thank you!

    Was this answer helpful?

  3. AI answer

    2025-12-13T09:46:36.8233333+00:00

    The error you are encountering while testing an agent in AI Foundry could be due to various reasons. Since the error is recurring, it may indicate a persistent issue with the agent or the environment. Here are some steps you can take to troubleshoot the problem:

    1. Retry the Request: As suggested in the error message, you can try to retry your request. Sometimes, transient issues can cause temporary errors.
    2. Check for Known Issues: Review the Azure status page or documentation for any known issues that might be affecting the AI Foundry service.
    3. Contact Azure Support: If the problem persists after multiple attempts, it is advisable to contact Azure support. Include the request IDs you shared (68ed3d2a-ffb0-4d1e-b608-b8cbf058df32 and 9e2f7fba-b4a5-45a5-8d6a-1efc077e31ac) in your support request for better assistance.
    4. Review Agent Configuration: Ensure that the agent is configured correctly and that there are no issues with the prompt or the model being used.
    5. Monitor Resource Usage: Sometimes, resource limitations can cause errors. Make sure that your Azure resources are not being overutilized.

    If the issue continues, detailed logs and error messages from the AI Foundry environment may provide additional insights into the root cause of the problem.

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.