Showing error 400 unsupported parameter max token after I changed the model to 5.1

Question

Showing error 400 unsupported parameter max token after I changed the model to 5.1

KWOK, Lon Lee (Admin) 0

I have a web app hosting Azure OpenAI gpt services. I have changed the model from gpt-4o to gpt-5.1 recently. However I received error 400-unsupported parameter, max token is not supported in this model.

I have changed the name to max-completion_tokens in the Environment Variables in the Web App services but still returned the same error when using the gpt services.

SRILAKSHMI C 14,815 Reputation points Microsoft External Staff Moderator

2026-03-04T05:52:05.79+00:00
Hello KWOK, Lon Lee (Admin),

Welcome to Microsoft's Q&A section and thank you for contacting us.

I understand that you switched your deployment to gpt-5.1, but your request is still using a parameter that this model does not accept.

The 400 – unsupported parameter error happens when the request body contains a field that GPT-5.1 doesn’t recognize.

What’s Happening

There are two common causes after moving from gpt-4o → gpt-5.1:

1.Wrong endpoint

GPT-5.1 must be called via the Chat Completions endpoint:

/openai/deployments/{deployment-name}/chat/completions?api-version=...

If you call the older Completions endpoint, you’ll receive unsupported parameter errors.

2.Incorrect token parameter

For GPT-5.x models in Azure OpenAI:

You must use:

"max_completion_tokens": 500

NOT:

max token

max-token

max-completion_tokens

max_tokens

If your payload still contains "max_tokens", GPT-5.1 will return 400 – unsupported parameter

Environment Variables Alone Don’t Fix It

Changing the name in App Service environment variables is not enough.

You must ensure your actual request body contains:

{

If your backend code is still mapping an environment variable to "max_tokens" in JSON, the error will persist.

SDK Examples

JavaScript

await client.chat.completions.create({

Python

client.chat.completions.create(

Also Verify

You are using the Chat Completions endpoin, You are not calling /completions , Your API version supports GPT-5.1, Your SDK version is up to date, The request body contains max_completion_tokens not max_token

I hope this helps, Do let me know if you have further queries.

Thank you!
SRILAKSHMI C 14,815 Reputation points Microsoft External Staff Moderator

2026-03-05T03:16:25.0066667+00:00

Hi KWOK, Lon Lee (Admin),

Did you get any chance to review the above response. Do let me know if you have any further queries.

Thank you!

1 answer

Your answer

SRILAKSHMI C 14,815 Reputation points Microsoft External Staff Moderator

2026-03-04T05:52:05.79+00:00

Hello KWOK, Lon Lee (Admin),

Welcome to Microsoft's Q&A section and thank you for contacting us.

I understand that you switched your deployment to gpt-5.1, but your request is still using a parameter that this model does not accept.

The 400 – unsupported parameter error happens when the request body contains a field that GPT-5.1 doesn’t recognize.

What’s Happening

There are two common causes after moving from gpt-4o → gpt-5.1:

1.Wrong endpoint

GPT-5.1 must be called via the Chat Completions endpoint:

/openai/deployments/{deployment-name}/chat/completions?api-version=...

If you call the older Completions endpoint, you’ll receive unsupported parameter errors.

2.Incorrect token parameter

For GPT-5.x models in Azure OpenAI:

You must use:

"max_completion_tokens": 500

NOT:

max token

max-token

max-completion_tokens

max_tokens

If your payload still contains "max_tokens", GPT-5.1 will return 400 – unsupported parameter

Environment Variables Alone Don’t Fix It

Changing the name in App Service environment variables is not enough.

You must ensure your actual request body contains:

{

If your backend code is still mapping an environment variable to "max_tokens" in JSON, the error will persist.

SDK Examples

JavaScript

await client.chat.completions.create({

Python

client.chat.completions.create(

Also Verify

You are using the Chat Completions endpoin, You are not calling /completions , Your API version supports GPT-5.1, Your SDK version is up to date, The request body contains max_completion_tokens not max_token

I hope this helps, Do let me know if you have further queries.

Thank you!
SRILAKSHMI C 14,815 Reputation points Microsoft External Staff Moderator

2026-03-05T03:16:25.0066667+00:00

Hi KWOK, Lon Lee (Admin),

Did you get any chance to review the above response. Do let me know if you have any further queries.

Thank you!

Answer 1

KWOK, Lon Lee (Admin) 0

Hi SRILAKSHMI C,

Thanks for you guide.

I have tried to follow your instruction but not able to solve the issue.

Previously I deployed GPT-4o from the portal to an Azure Web services. It seems like the only place where I can change the max_completion_token is at the Environment Variables in that Web App services. I tried to deploy GPT-5.1 to another web app from the portal but that option does not display. However if I change the model to GPT-4o, the deploy option will appear.

Therefore I'm not sure whether is it there is an issue on model GPT-5 and above to be deployed to Azure Web services.

SRILAKSHMI C 14,815 Reputation points Microsoft External Staff Moderator

2026-03-06T05:41:00.0733333+00:00
Hi KWOK, Lon Lee (Admin),

Thank you for the update and for trying the suggested steps.

The behavior you are seeing is expected and is not related to a limitation with deploying GPT-5.1 to Azure Web Apps.

Clarification on Deployment

Azure OpenAI models such as GPT-4o or GPT-5.1 are not deployed to Azure Web App directly. Instead, they are deployed to an Azure OpenAI resource, and your Web App simply calls the deployed model through the API.

So when you previously selected Deploy to Web App from the portal, that template was created specifically for certain models (such as GPT-4o). Newer models like GPT-5.x may not appear in that quick-deploy template, but they can still be used normally from a Web App by calling the Azure OpenAI endpoint.

Why the Error Is Still Occurring

The 400 – unsupported parameter: max token error indicates that the request payload sent from your application still contains an unsupported parameter.

Updating the Environment Variable name in the Web App does not automatically change the JSON request body that your application sends to the Azure OpenAI API.

For GPT-5.x models, the request must use:

"max_completion_tokens": <value>

and must not contain:

max_tokens

If the application code is still constructing the request with max_tokens, the API will continue to return the 400 error even if the environment variable name has been updated.

Recommended Step

Please check the actual request payload generated by your application code (for example in the backend service or API call) and confirm that:

The endpoint being called is:

/openai/deployments/{deployment-name}/chat/completions

The request body includes:

"max_completion_tokens": <value>

The request body does not include max_tokens anywhere.

Thank you!

Share via

Showing error 400 unsupported parameter max token after I changed the model to 5.1

1 answer

Your answer