o3

Was that a pretty long time ago? Because I use o3 daily and have never hit a limit. The only "limit" I have really with o3 is my patience -- it's a slow model for regular use. If I don't need its intelligence I'll use o4-mini or even 4o for speed, saving o3 for the prompts that really need it. These days the only model I find I get rate-limited on frustratingly quickly is 4.5. It's clear OpenAI does not really want you using that one, despite the fact that's very good (and probably very expensive for them)! Pretty underrated imo.

I think I saw o3 hallucinate more in a single day of heavy use then I saw o3-mini do in a month. That said: so far, I'm putting up with it because o3 is smart.

OpenAI o3 and o4-mini are massive disappointments for me so far. I have a private benchmark of 10 proof-based geometric group theory questions that I throw at new models upon release. Both new models gave inconsistent answers, always with wrong or fake proofs, or using assumptions that are not in the queation, and are often outright unsatisfiable. The now inaccessible o3-mini was not great, but much better than o3 and o4-mini at these questions: o3-mini can give approximately correct proof sketches for half of them, whereas I can't get a single correct proof sketch out of o3 full. o4-mini performs slightly worse than o3-mini. I think the allegations that OpenAI cheated FrontierMath have unambiguously been proven correct by this release.