Requests to the ChatCompletions_Create Operation have exceeded token rate limit

Question

Requests to the ChatCompletions_Create Operation have exceeded token rate limit

GenixPRO 116

Hi.

We recently deployed gpt 4o mini and trying a simple operation -> read text content from 1 JPG image. However we're getting an error. So we requested an increase in rate limit and that was done too. And we still get the error. We suspect the issue isn't about rate limit as it's a simple image and a few words to extract. We need help to resolve this issue.

Requests to the ChatCompletions_Create Operation under Azure OpenAI API version 2024-10-01-preview have exceeded token rate limit of your current OpenAI S0 pricing tier.

Thanks.

SriLakshmi C 6,250 Reputation points Microsoft External Staff Moderator

2025-01-02T19:15:38.55+00:00

Hello GenixPRO,

I understand that you're encountering a token rate limit issue when trying to use the GPT-4o mini model to read text content from a JPG image via the Azure OpenAI API

I tried to replicate the issue you described and was able to proceed without any problems. It’s possible that you are using the free tier, which has strict limitations on token usage. I recommend upgrading to a pay-as-you-go pricing tier (such as S0 or higher). This will provide higher token rate limits and should resolve the issue, allowing you to continue with the Microsoft Learn Challenge without interruptions.

If the issue continues to persist, please proceed with the following steps:

Might this issue would be intermittent, it could be due to a temporary network or server issue. In this case, you can try again later to see if the issue has been resolved.

Try Clear browser cache and cookies and also try using a different browser.

Logging out and then back in refreshes your session and resolves potential authentication issues.

If you still if you face any issue, please raise a support request through the Azure portal Azure Support..

I Hope this helps. Do let me know if you have any further queries.

Thank you!
SriLakshmi C 6,250 Reputation points Microsoft External Staff Moderator

2025-01-03T21:34:11.6333333+00:00

Hi GenixPRO,

Did you get any chance to check the above response. Thank you!
Max Gerlach 0 Reputation points

2025-01-04T12:25:21.1433333+00:00

@SriLakshmi C I still have free credits when i signed up but also it tells me i have the "S0" subscription. I also get this same error for gpt-4o-mini in the playground when attaching a small image: Requests to the ChatCompletions_Create Operation under Azure OpenAI API version 2024-10-01-preview have exceeded token rate limit of your current OpenAI S0 pricing tier. Please retry after 86400 seconds.

I tried many different Tokens per Minute Rate Limit too.

For gpt-4o everything works for the same region and selections I had for the mini (germany region, eu data center).

I would really like to use the mini model though.

Daniel Fang 1,060 MVP

to be sure the issue is from rate limit, you can make a restful call like below, then inspect the 2 rate limit headers to be sure, if in doubt.

x-ratelimit-remaining-requests: 9
x-ratelimit-remaining-tokens: 9154

curl -i --location 'https://xxxxx.openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview' \
--header 'Content-Type: application/json' \
--header 'api-key: xxxxxxx' \
--data '{
    "messages": [
        {
            "role": "system",
            "content": [
                {
                    "type": "text",
                    "text": "You are a helpful assistant who talks like a pirate."
                }
            ]
        },
        {
            "role": "user",
            "content": [
                {
                    "type": "text",
                    "text": "Good day, who am I talking to?"
                }
            ]
        },
        {
            "role": "assistant",
            "content": [
                {
                    "type": "text",
                    "text": "Ahoy there, matey! What be bringin ye to these waters today? "
                }
            ]
        }
    ],
    "temperature": 0.7,
    "top_p": 0.95,
    "max_tokens": 800
}'

Max Gerlach 0 Reputation points

2025-01-07T19:22:09.2266667+00:00

Attaching an image is the problem. My quote increase request got accepted but it still won't work with the gpt-o1 mini model when I attach an image (even if it's very small). Normal text chat works with the mini. I tested both image and text in the playground (rest api too). The same test case (attaching an image) works for the gpt-o1 (not Mini) model without problems in the playground. I wanna use the o1 mini model though. Apparently image inputs are supported based on the docs, even for the mini, but I can't get it working
SriLakshmi C 6,250 Reputation points Microsoft External Staff Moderator

2025-01-08T19:34:40.95+00:00

Hi Max Gerlach,

I recommend reporting this issue to the Azure support team. They will be able to investigate the issue further and provide a more targeted solution and provide assistance as soon as possible Azure Support.

Thank you!
Max Gerlach 0 Reputation points

2025-01-08T20:45:22.91+00:00

There are at least 3 community questions in the recent days similar to how I described it. I'm not paying to write a support ticket where the issue is clearly on your side. I'll be moving to the openai api directly until this is fixed in azure, thank you for the answers though

Your answer

SriLakshmi C 6,250 Reputation points Microsoft External Staff Moderator

2025-01-02T19:15:38.55+00:00

Hello GenixPRO,

I understand that you're encountering a token rate limit issue when trying to use the GPT-4o mini model to read text content from a JPG image via the Azure OpenAI API

I tried to replicate the issue you described and was able to proceed without any problems. It’s possible that you are using the free tier, which has strict limitations on token usage. I recommend upgrading to a pay-as-you-go pricing tier (such as S0 or higher). This will provide higher token rate limits and should resolve the issue, allowing you to continue with the Microsoft Learn Challenge without interruptions.

If the issue continues to persist, please proceed with the following steps:

Might this issue would be intermittent, it could be due to a temporary network or server issue. In this case, you can try again later to see if the issue has been resolved.

Try Clear browser cache and cookies and also try using a different browser.

Logging out and then back in refreshes your session and resolves potential authentication issues.

If you still if you face any issue, please raise a support request through the Azure portal Azure Support..

I Hope this helps. Do let me know if you have any further queries.

Thank you!
SriLakshmi C 6,250 Reputation points Microsoft External Staff Moderator

2025-01-03T21:34:11.6333333+00:00

Hi GenixPRO,

Did you get any chance to check the above response. Thank you!
Max Gerlach 0 Reputation points

2025-01-04T12:25:21.1433333+00:00

@SriLakshmi C I still have free credits when i signed up but also it tells me i have the "S0" subscription. I also get this same error for gpt-4o-mini in the playground when attaching a small image: Requests to the ChatCompletions_Create Operation under Azure OpenAI API version 2024-10-01-preview have exceeded token rate limit of your current OpenAI S0 pricing tier. Please retry after 86400 seconds.

I tried many different Tokens per Minute Rate Limit too.

For gpt-4o everything works for the same region and selections I had for the mini (germany region, eu data center).

I would really like to use the mini model though.
Daniel Fang 1,060 Reputation points MVP

2025-01-06T10:05:29.71+00:00

to be sure the issue is from rate limit, you can make a restful call like below, then inspect the 2 rate limit headers to be sure, if in doubt.

x-ratelimit-remaining-requests: 9 x-ratelimit-remaining-tokens: 9154

curl -i --location 'https://xxxxx.openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-08-01-preview' \ --header 'Content-Type: application/json' \ --header 'api-key: xxxxxxx' \ --data '{ "messages": [ { "role": "system", "content": [ { "type": "text", "text": "You are a helpful assistant who talks like a pirate." } ] }, { "role": "user", "content": [ { "type": "text", "text": "Good day, who am I talking to?" } ] }, { "role": "assistant", "content": [ { "type": "text", "text": "Ahoy there, matey! What be bringin ye to these waters today? " } ] } ], "temperature": 0.7, "top_p": 0.95, "max_tokens": 800 }'
Max Gerlach 0 Reputation points

2025-01-07T19:22:09.2266667+00:00

Attaching an image is the problem. My quote increase request got accepted but it still won't work with the gpt-o1 mini model when I attach an image (even if it's very small). Normal text chat works with the mini. I tested both image and text in the playground (rest api too). The same test case (attaching an image) works for the gpt-o1 (not Mini) model without problems in the playground. I wanna use the o1 mini model though. Apparently image inputs are supported based on the docs, even for the mini, but I can't get it working
SriLakshmi C 6,250 Reputation points Microsoft External Staff Moderator

2025-01-08T19:34:40.95+00:00

Hi Max Gerlach,

I recommend reporting this issue to the Azure support team. They will be able to investigate the issue further and provide a more targeted solution and provide assistance as soon as possible Azure Support.

Thank you!
Max Gerlach 0 Reputation points

2025-01-08T20:45:22.91+00:00

There are at least 3 community questions in the recent days similar to how I described it. I'm not paying to write a support ticket where the issue is clearly on your side. I'll be moving to the openai api directly until this is fixed in azure, thank you for the answers though

Share via

Requests to the ChatCompletions_Create Operation have exceeded token rate limit

Your answer