OpenAI has launched a more powerful version of its o1 โreasoningโ AI model, o1-pro, in its developer API.
According to OpenAI, o1-pro uses more computing than o1 to provide โconsistently better responses.โ Currently, itโs only available to select developers โ those whoโve spent at least $5 on OpenAI API services โ and itโs pricey. Very pricey.
OpenAI is charging $150 per million tokens (~750,000 words) fed into the model and $600 per million tokens generated by the model. Thatโs twice the price of OpenAIโs GPT-4.5 for input and 10x the price of regular o1.
OpenAI is betting that o1-proโs improved performance will convince developers to pay those princely sums.
โO1-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,โ an OpenAI spokesperson told TechCrunch. โAfter getting many requests from our developer community, weโre excited to bring it to the API to offer even more reliable responses.โ
Yet early impressions of o1-pro, which has been available in OpenAIโs AI-powered chatbot platform ChatGPT for ChatGPT Pro subscribers since December, werenโt incredibly positive. The model struggled with Sudoku puzzles, users found, and was tripped up by simple optical illusion jokes.
Furthermore, certain OpenAI internal benchmarks from late last year showed that o1-pro performed only slightly better than the standard o1 on coding and math problems. It did answer those problems more reliably, however, the benchmarks found.
