![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(233.6, 77%, 32%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ED%3C/text%3E%3C/svg%3E)
Azure AI Language
An Azure service that provides natural language capabilities including sentiment analysis, entity extraction, and automated question answering.
Hi @DMcCrea,
That 401 is almost always a routing or authentication mismatch. In your call you’re mixing the Azure OpenAI style route (/openai/deployments/...) with an Azure AI Foundry Models resource and key. Foundry resources use a different base URL:
https://<your-resource>.services.ai.azure.com/models
and the chat-completions route is simply:
POST https://<your-resource>.services.ai.azure.com/models/chat/completions?api-version=2024-05-01
What to change in Postman
| Setting | Correct value | Why | |
|---|---|---|---|
| URL | https://<resource>.services.ai.azure.com/models/chat/completions?api-version=2024-05-01 | Foundry’s Model Inference endpoint(learn.microsoft.com) | |
| URL | https://<resource>.services.ai.azure.com/models/chat/completions?api-version=2024-05-01 |
Foundry’s Model Inference endpoint([learn.microsoft.com](https://learn.microsoft.com/en-us/azure/ai-foundry/model-inference/concepts/endpoints"Endpoints for Azure AI Foundry Models - Azure AI Foundry | Microsoft Learn")) |
| Header | api-key: <KEY1 or KEY2>Content-Type: application/json |
Foundry accepts the same key header as Azure OpenAI([learn.microsoft.com](https://learn.microsoft.com/en-us/azure/ai-foundry/model-inference/concepts/endpoints"Endpoints for Azure AI Foundry Models - Azure AI Foundry | Microsoft Learn")) |
| Body | json { "model":"gpt-4o-mini", "messages":[{ "role":"user","content":"Hello!"}] } |
model should match the deployment name you see in the portal |
A minimal curl that should return 200:
curl -X POST \
"https://myfoundry.services.ai.azure.com/models/chat/completions?api-version=2024-05-01" \
-H "api-key: <YOUR_KEY>" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Hello"}]}'
Calling an Agent instead of a raw model?
If you actually want to hit an Agent Service project (threads / assistants flow) you switch to AAD tokens, not keys, and the endpoint shape changes again:
https://<resource>.services.ai.azure.com/api/projects/<project-name>/...
Authorization: Bearer <token from `az account get-access-token --resource https://ai.azure.com`>
Full token/endpoint example is in the agent quick-start docs.
Troubleshooting checklist
Double-check you copied the full key with no trailing spaces and that the key hasn’t been regenerated.
Make sure the resource name in the host matches the one that generated the key (region must match too).
Verify the deployment name (model property in the body) exactly matches what you see in Foundry ▶ Deployments.
Use the GA API version 2024-05-01 unless you specifically need a preview feature.
With the correct endpoint + header your request should authenticate and return 200. Let me know if you are still facing a problem.
Best Regards,
Jerald Felix