Failed to call Azure OpenAI API: { "statusCode": 401, "message": "Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com), or have expired." }

Question

Failed to call Azure OpenAI API: { "statusCode": 401, "message": "Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com), or have expired." }

Ravi Modukuri 15

Chandra Boorla 14,510 Reputation points Microsoft External Staff Moderator

2024-09-13T18:40:57.6366667+00:00

Hi @Ravi Modukuri

Greetings & Welcome to Microsoft Q&A forum! Thanks for posting your query!

Apologies for the issue that you are facing.

The error message you are seeing indicates an issue with the authentication process when trying to call the Azure OpenAI API.

This could happen for several reasons, such as your API key is missing or invalid, or the endpoint is incorrect.

Please make sure that you have a valid API key and endpoint for GPT-4o.

Also, make sure that you are using latest and updated python versions, Openai and api_version.

See the similar issues below which has been addressed and let me know if that helps.

https://learn.microsoft.com/en-us/answers/questions/1685378/how-to-use-gpt-4o-without-azure-cognitive-search?page=0&orderby=Helpful#answers.

Failed to call Azure OpenAI API: failed with error:invalid_api_key.

I hope this information helps. Please do let us know if you have any further queries.
Chandra Boorla 14,510 Reputation points Microsoft External Staff Moderator

2024-09-16T17:20:34.99+00:00

Hi @Ravi Modukuri

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Chandra Boorla 14,510 Reputation points Microsoft External Staff Moderator

2024-09-17T21:28:02.3266667+00:00

Hi @Ravi Modukuri

Just checking in to see on the last response. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.

Thank you.
Teja Arendala 0 Reputation points

2024-11-14T12:01:31.02+00:00

Hi @Ravi Modukuri

I'm also facing a similar issue while creating an assistant through Python code. FYI, I have the latest versions of Open AI and Python. I tried it in Postman, and it works there with same api-key,api-version and endpoint.
Geena Kim 0 Reputation points

2024-11-16T02:35:12.5466667+00:00

Can someone give an example? I matched the api version the same as the model deployment endpoint's target uri- for example, (Note that the real address is masked with [azureopenai-resourcename] since everyone's case will be different) https://[azureopenai-resourcename].openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview

Then I set the required environment variables like

export AZURE_OPENAI_API_KEY="<replaced with my key>"

export AZURE_OPENAI_ENDPOINT="https://[azureopenai-resourcename].openai.azure.com/"

export OPENAI_API_TYPE="azure"

export OPENAI_API_VERSION="2024-08-01-preview" ## This is to match the api version that shows up in the target uri. But it still gives the error messages like:

INFO:httpx:HTTP Request: POST https://[azureopenai-resourcename].openai.azure.com//openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview "HTTP/1.1 401 Unauthorized"

Error code: 401 - {'statusCode': 401, 'message': 'Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com), or have expired.'}

When I use LangChain to load embedding model, it works ok, but while using reranker using gpt model(s) it makes a problem. I tested with export OPENAI_API_TYPE="openai" and used my personal account (in this case it didn't need any other auth other than OPENAI_API_KEY. So it seems the langchain's rerank works and somehow it breaks when azure credentials are passed. So other than those envs (AZURE_OPENAI_API_KEY, AZURE_OPENAI_ENDPOINT, OPENAI_API_TYPE, OPENAI_API_VERSION), what "Access token" is required?? And what is https://cognitiveservices.azure.com?? Do we need to set up another endpoint other than the model deployment (as above uri shows)?

2 answers

Your answer

Chandra Boorla 14,510 Reputation points Microsoft External Staff Moderator

2024-09-13T18:40:57.6366667+00:00

Hi @Ravi Modukuri

Greetings & Welcome to Microsoft Q&A forum! Thanks for posting your query!

Apologies for the issue that you are facing.

The error message you are seeing indicates an issue with the authentication process when trying to call the Azure OpenAI API.

This could happen for several reasons, such as your API key is missing or invalid, or the endpoint is incorrect.

Please make sure that you have a valid API key and endpoint for GPT-4o.

Also, make sure that you are using latest and updated python versions, Openai and api_version.

See the similar issues below which has been addressed and let me know if that helps.

https://learn.microsoft.com/en-us/answers/questions/1685378/how-to-use-gpt-4o-without-azure-cognitive-search?page=0&orderby=Helpful#answers.

Failed to call Azure OpenAI API: failed with error:invalid_api_key.

I hope this information helps. Please do let us know if you have any further queries.
Chandra Boorla 14,510 Reputation points Microsoft External Staff Moderator

2024-09-16T17:20:34.99+00:00

Hi @Ravi Modukuri

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Chandra Boorla 14,510 Reputation points Microsoft External Staff Moderator

2024-09-17T21:28:02.3266667+00:00

Hi @Ravi Modukuri

Just checking in to see on the last response. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.

Thank you.
Teja Arendala 0 Reputation points

2024-11-14T12:01:31.02+00:00

Hi @Ravi Modukuri

I'm also facing a similar issue while creating an assistant through Python code. FYI, I have the latest versions of Open AI and Python. I tried it in Postman, and it works there with same api-key,api-version and endpoint.
Geena Kim 0 Reputation points

2024-11-16T02:35:12.5466667+00:00

Can someone give an example? I matched the api version the same as the model deployment endpoint's target uri- for example, (Note that the real address is masked with [azureopenai-resourcename] since everyone's case will be different) https://[azureopenai-resourcename].openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview

Then I set the required environment variables like

export AZURE_OPENAI_API_KEY="<replaced with my key>"

export AZURE_OPENAI_ENDPOINT="https://[azureopenai-resourcename].openai.azure.com/"

export OPENAI_API_TYPE="azure"

export OPENAI_API_VERSION="2024-08-01-preview" ## This is to match the api version that shows up in the target uri. But it still gives the error messages like:

INFO:httpx:HTTP Request: POST https://[azureopenai-resourcename].openai.azure.com//openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview "HTTP/1.1 401 Unauthorized"

Error code: 401 - {'statusCode': 401, 'message': 'Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com), or have expired.'}

When I use LangChain to load embedding model, it works ok, but while using reranker using gpt model(s) it makes a problem. I tested with export OPENAI_API_TYPE="openai" and used my personal account (in this case it didn't need any other auth other than OPENAI_API_KEY. So it seems the langchain's rerank works and somehow it breaks when azure credentials are passed. So other than those envs (AZURE_OPENAI_API_KEY, AZURE_OPENAI_ENDPOINT, OPENAI_API_TYPE, OPENAI_API_VERSION), what "Access token" is required?? And what is https://cognitiveservices.azure.com?? Do we need to set up another endpoint other than the model deployment (as above uri shows)?

Answer 1

Daniel Fang 1,060 MVP

It could also be a delay for the service to be ready and servicing requests if it is a new resource you just created. Checking all the URLs and API keys and try after 1h. The other thing is that also try the secondary key, there is once that primary key not working for us but secondary key actually worked.

Answer 2

Can someone give an example? I matched the api version the same as the model deployment endpoint's target uri- for example, (Note that the real address is masked with [azureopenai-resourcename] since everyone's case will be different)
https://[azureopenai-resourcename].openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview

Then I set the required environment variables like

export AZURE_OPENAI_API_KEY="<replaced with my key>"

export AZURE_OPENAI_ENDPOINT="https://[azureopenai-resourcename].openai.azure.com/"

export OPENAI_API_TYPE="azure"

export OPENAI_API_VERSION="2024-08-01-preview" ## This is to match the api version that shows up in the target uri.

But it still gives the error messages like:

INFO:httpx:HTTP Request: POST https://[azureopenai-resourcename].openai.azure.com//openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview "HTTP/1.1 401 Unauthorized"

Error code: 401 - {'statusCode': 401, 'message': 'Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com), or have expired.'}

When I use LangChain to load embedding model, it works ok, but while using reranker using gpt model(s) it makes a problem.
I tested with export OPENAI_API_TYPE="openai" and used my personal account (in this case it didn't need any other auth other than OPENAI_API_KEY.
So it seems the langchain's rerank works and somehow it breaks when azure credentials are passed.

So other than those envs (AZURE_OPENAI_API_KEY, AZURE_OPENAI_ENDPOINT, OPENAI_API_TYPE, OPENAI_API_VERSION), what "Access token" is required??
And what is https://cognitiveservices.azure.com??
Do we need to set up another endpoint other than the model deployment (as above uri shows)?

JayAmir-0051 0 Reputation points

2024-12-13T23:25:59.6833333+00:00

THis is two API call. First we will get the token and then pass the token in the second API call.

Share via

Failed to call Azure OpenAI API: { "statusCode": 401, "message": "Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com), or have expired." }

2 answers

Your answer