Hacker News Reader

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

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