I was interacting and conversing with the o3-mini model after enabling the “reason” feature. Once I had exhausted the allowed limit for this feature, the model automatically switched to GPT-4o, but I was no longer able to activate the reasoning capability. During this transition, I began noticing recurring bugs in the responses. The model provided answers to questions I had asked days ago instead of responding to my most recent inquiry. This issue occurred multiple times throughout that session