pricing on the new o1 models is not unreasonable if you’ve ever tried agentic/chain-of-thought reasoning with previous models. the repeated calls of the latter make the former quite attractive.