[Max-q]

I would say using performance of Perplexity as a benchmark for the quality of o1-pro is a stretch?

[yosito]

Find third party benchmarks of the relevant models and then this discussion is worth having. Otherwise, it's just speculation.