Hajra Naz
OpenAI has launched O1-Pro, a more advanced version of its O1 “reasoning” AI model, now available through its developer API.
Enhanced Performance, Higher Costs
According to OpenAI, O1-Pro uses more computing power than O1 to deliver consistently better responses. However, access is limited to select developers—only those who have spent at least $5 on OpenAI API services can use it.
And it comes at a hefty price. OpenAI is charging:
- $150 per million tokens (~750,000 words) for input
- $600 per million tokens for output
That’s twice the cost of GPT-4.5 for input and 10x the price of standard O1.
Justifying the Price Tag
OpenAI believes O1-Pro’s improved reasoning abilities will convince developers to pay the premium. An OpenAI spokesperson told.
“O1-Pro in the API is a version of O1 that uses more computing to think harder and provide even better answers to the hardest problems,”. After numerous requests from developers, we’re excited to bring it to the API to offer more reliable responses.”
Mixed Early Impressions
Despite its promise, early feedback on O1-Pro has been lukewarm. Since December, it has been available to ChatGPT Pro subscribers, but users noted struggles with Sudoku puzzles and difficulty understanding optical illusion jokes.
Moreover, OpenAI’s internal benchmarks suggest O1-Pro performs only slightly better than standard O1 in coding and math problems. While it does provide more consistent responses, the overall improvement appears marginal.
