OpenAI has launched o1-pro, an upgraded version of its reasoning model, o1. This enhanced model promises improved performance but comes with a significantly higher price tag.
OpenAI’s announcement highlights o1-pro’s increased computational power, leading to “consistently better responses.” Key features include vision support, function calling, structured outputs, and compatibility with the Responses and Batch APIs. However, this enhanced performance comes at a cost: $150 per 1 million input tokens and $600 per 1 million output tokens. As reported by TechCrunch, this pricing structure positions o1-pro at double the cost of GPT-4.5 and ten times the cost of the baseline o1 model.
O1-Pro: Designed for Developers
OpenAI explicitly targets o1-pro towards developers rather than general users. An OpenAI spokesperson told TechCrunch that o1-pro utilizes more computing power to “think harder” and deliver superior responses to complex problems. The company developed o1-pro in response to developer community demand for more reliable responses within the API.
Access to o1-pro is currently limited to select developers on tiers 1-5 based on their past spending on OpenAI’s API services. Higher-tier developers enjoy greater request allowances within specified timeframes.
Cost vs. Benefit: A Developer Dilemma
The high cost of o1-pro raises questions about developer adoption. Following its introduction in ChatGPT Pro a few months prior, user feedback was mixed.
Some Reddit users criticized o1-pro’s real-world performance, deeming it “pathetic” despite strong benchmark results. However, others praised its programming capabilities, particularly when provided with detailed instructions.
Availability and Access
Developers willing to invest in o1-pro can access it through OpenAI’s development platform. Whether the enhanced performance justifies the substantial cost remains to be seen.
