This model was released before the ID verification scheme was used to deny new models and features. Instead, there is an undisclosed multi-layer “trust tier” of organization metrics (like age and usage) being applied to you that is/was also used to deny some organizations access to models, like even o3-mini.
My observation:
Dark patterns:
If you are at the requisite payment tier, you can send an email to support@openai.com with your org ID from your org login email, and request that the model be provisioned for your organization. You’ll likely need to spell out what you’re talking about explicitly in order for the request to be understood by the first support tier and their default of running a “deny performing anything useful” AI on your message - clarifying what the model is and what they need to do (to escalate to someone who can actually change the models your org has access to).