Tier 5 here without access. I have a use case for this that I’m (impatiently) waiting for, would appreciate getting access sooner than later!
They said a few weeks when launched. That would point to tomorrow as the deadline on the original tweet. A few = 3 the last time I checked. but it still makes me wonder if o3-mini which is supposed to launch this month and be better and cheaper will take forever to roll out too. I wonder if all other future releases will follow this same BAD role out plan where 99% of us have to wait weeks or longer to get access. This is one thing OpenAI needs to vastly improve on. When you announce something new and is said to be available ANY paying API customer should be able to access right away even with very small rate limits. Even as a pro user I should get access. I have access to the model anytime I need it in a chat interface even pro mode. but still don’t have API access.
Still no access for me. I sent an email to OpenAI yesterday and I got this response:
" We apologize for any inconvenience. You should already have access to the o1 models. If access is still unavailable, it may be due to account settings or organizational alignment. Steps to resolve the issue include:
-
Log Out and Log Back In: Please log out and then log back in to confirm you are using the account associated with Tier 5 access.
-
Verify Your Organization: Ensure access to the correct organization linked with Tier 5. To switch organizations on the OpenAI API platform, follow these steps:
- Hover over the organization’s name in the top-left corner of the page.
- Select the desired organization from the appearing list.
For detailed instructions, refer to our article on “Managing your work in the API platform with Projects.”"
But I’ve already checked this (multiple times) and don’t have access. So I’m thinking the member of staff who responded misspoke, or thought I meant I don’t have o1-mini and preview. Still holding out hope I’ll have access soon. I can use it in my ChatGPT Pro subscription, but I need to be able to use it in the API for my use cases.
This is the last update that I’ve had:
“We apologize for the confusion in our previous response. Due to an issue with the rollout, access to the full o1 model has been temporarily withdrawn. Our team is actively working to resolve this, and updates will be provided once the rollout resumes.”
I’m not sure I’ve read any updates about OpenAI halting the rollout and it would be good to understand when this is likely to resume if this is the case.
Yeah the communication from OpenAI in regard to this rollout has been horrible. That’s one thing they need to improve upon. Rollout of models in general has never been this bad before as normally I could use another provider until I got access but this time none of those providers even have access. My guess this has something to do with the new Pro tier and the surprise usage on that. That is only my guess.
Rollout has not been paused. (Not sure why our support team member responded with that, but we’ll address.)
We didn’t add more users during Christmas/New Year’s, but we’re continuing the rollout of o1 in the API. The plan remains unchanged (and has no relation to ChatGPT Pro mode) — the transition from o1-preview to o1 will take weeks, and we’re adding more developers in (randomized) batches over time. You’ll receive an email once you have access. Again, we’re sorry for making you wait, but rollout to tier 5 will complete soon!
Thanks finally for an update and clarification. The term may not have come from a “support team member”, but an observer:
Perhaps not paused - in the same way my water is not technically “paused” when I use the main shutoff valve when going on vacation… 
More diplomatically, someone paid to write nice messages might give you a free communique to round out this topic:
Our configuration management team did not roll out o1 automatically to any new tier-5 users during the holiday break, to ensure system stability during this period of reduced staffing. We apologize if this did not meet expectations of many that anticipated arrival of this new model shortly after the announcement, and we will try to communicate our plans more clearly in the future. We hope you will enjoy discovering applications for this new series of reasoning models as we roll it out to more of our valued organizations in the coming weeks, in a manner that still ensures a high quality of service you expect.
It was from a support team member. I copied the email reply directly above. I’m not sure why they said it. They must have misspoke again.
It would be nice to get an idea of how long this will take, this does concern me with future releases as well. The communication hasn’t been great and it concerns me that I might be months behind in the future. Hopefully this is a one off issue.
Thanks for the update. Has the support for developer message been rolled out as well?
Using the snippet in the Reasoning documentation works fine. But if I add another message with the role developer before the user message, I receive this error:
messages[0].role' does not support 'developer' with this model.
Any reason this is happening? It gives the same response if I set the role as system. There are no examples in the documentation as well with role=developer. Please give some clarity on this.
Same problem. Also I can’t ask a visual question. I can ask only text question without role (system/developer)
Tier 5 here, same problem, still have no access to o1
Paid my way to tier 5 but now I guess I still have to play the waiting game … Already seen some promising results with model 4o however my application needs something just a bit smarter. Very excited to start testing when I get acess.
Pushed my sub to tier 5 to get o1 access and I’m sorry to say it is not available. Any chance you can tell me when I will be able to use it as advertised?
It will be with you soon…