Agents and Models and and and
I have been an OpenAI customer since the day they opened up to the public. Chat and API with a fair bit of use. I have been using Codex Oauth without a problem but Codex is really wearing my patience and want to try Claude. I have it connected and routed to my coding agency and my first request hit 53% session limit on the Pro account.
I dropped from Opus to Sonnett at that point made a repeat request to test between the two and that took me to 76% on the 2nd request. Only 53% difference in models. I then asked with Sonnett to complete the work from the first evaluation request and it completed it in another 23%.
How are you managing this? and does the pro account lack that much that I should go to the 5x plan now?