Unable to use o1 via API, despite Paying For Pro and adding a key!

Discussion
1

Hi everyone,

I’m reaching out because I’ve been experiencing much frustration with my o1 API use or lack of it, and I’m hoping someone here can offer guidance or a solution.

I pay for Pro and have set my monthly spending limit appropriately. However, it’s been over a week, and I’m still unable to use the o1 API. I’ve tried multiple times to resolve the issue on my own, including:
1. Changing API keys multiple times.
2. Double-checking my spending limits and billing settings.

I continue to encounter the same issue for a paid service—especially one that will push what it can do.

I want it to work, especially since I’m paying for it. The whole point of paying for Pro is smooth access without these roadblocks.

If anyone from support or the community has any suggestions, I’d appreciate it. I’m happy to provide any additional details to help resolve this.

I appreciate any help you can provide.

Best regards,

3 Likes
2

Hey, we’re sorry you’re experiencing issues with your API key. Could you share some details? Are you receiving any error messages?

3

@deanrie

Please check your API key settings and try again. API Error:

{“error”:{“message”:“The model o1 does not exist or you do not have access to it.”,“type”:“invalid_request_error”,“param”:null,“code”:“model_not_found”}}
Request ID: da2ebe90-7b72-44d8-88d1-b354fc7fbcd1

I have changed it multiple times and updated all the software. Also, my Cluade hits a limit fast now.

thank you for your help

4

@deanrie If you have time and can help I would really like to talk to you about it

5

use Claude…

6

really

7

Thank you for your suggestion. However, I’m explicitly experiencing issues with using o1 and hoped to find a solution within this platform. Please provide more detailed assistance or direct me to the appropriate resources to resolve this issue.

1 Like
8

Have you verified that your OpenAI account has access to o1?

As of writing, o1 is only available to Tier 5 users via the API.

You can check yours in Settings → Limits, or directly at https://platform.openai.com/settings/organization/limits

If you can’t see o1 listed there, that’s likely your problem.

2 Likes
9
  1. There are multiple things that have to happen for this to work
  2. You have to have a key from openAI
  3. that key has to have been authorized to hit o1 by OpenAI. You can check this by doing: ```curl https://api.openai.com/v1/models \ -H “Authorization: Bearer $OPENAI_API_KEY” from a unix-like shell, or the equivalent in powershell. If OpenAI has not given it to you, there is nothing Cursor can do about it.
  4. Then you have to put the key into the config so that cursor can use it.

Upthread people claim that o1 is avaiable to people with level 5 tier with OpenAI. However, I am that level (from a bunch of Dall-e 3 requests) but do not have o1 API yet, so that’s not completely true.)

In your case I’d guess you don’t have the authorization from OpenAI to use it, which is why you don’t have it.

Final issue: it’s very confusing how Cursor lets users add random strings in the model list - this just opens the door for being confused about whether we should add ‘o1’ or ‘gpt-o1’ etc and how what you enter there is actually being mapped.

Seems like every single Cursor user must be constantly going to the models list page and trying to update the program to get the right model to show up.

Tl;DR You probably don’t have the authorization from OpenAI to hit this endpoint, so there’s nothing Cursor can do about it.

10

See this thread on the openai forums: it appears that tier5 is not sufficient to get access and that rollout has been paused for a while now:

etc

1 Like
11

As mentioned in these threads in @TPZone’s comments above, a tier 5 OpenAI account is not yet sufficient to access o1 via the API, and that rollout has been paused for a while now. As soon as you have o1 enabled on your account, your key should start working in Cursor.

1 Like