OpenAI has taken a significant step forward in the realm of artificial intelligence with the launch of o1-pro, a more powerful version of its "reasoning" AI model, o1. The new model is now available to select developers through OpenAI's developer API, but it comes with a hefty price tag that may limit its adoption.

According to OpenAI, o1-pro uses more computing power than its predecessor to provide "consistently better responses." This increased processing power is expected to result in more accurate and reliable answers to complex problems. However, this enhanced performance comes at a cost, with OpenAI charging $150 per million tokens fed into the model and $600 per million tokens generated by the model. This pricing structure is significantly higher than OpenAI's GPT-4.5, with input costs doubling and output costs increasing by a factor of 10 compared to the standard o1 model.

The exclusivity of o1-pro is another notable aspect of its launch. Currently, only developers who have spent at least $5 on OpenAI API services are eligible to access the new model. This limited availability may restrict the adoption of o1-pro, especially among smaller developers or startups with limited budgets.

OpenAI's decision to launch o1-pro is a response to requests from its developer community, who have been seeking more reliable and accurate AI models. "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," an OpenAI spokesperson explained to TechCrunch. Despite the high cost, OpenAI is betting that the improved performance of o1-pro will convince developers to pay the premium.

However, early impressions of o1-pro, which has been available to ChatGPT Pro subscribers since December, have been mixed. Users have reported that the model struggles with Sudoku puzzles and is easily tripped up by simple optical illusion jokes. Additionally, internal benchmarks from late last year showed that o1-pro performed only slightly better than the standard o1 on coding and math problems, although it did answer questions more reliably.

The launch of o1-pro raises important questions about the future of AI development and the role of cost in shaping the adoption of advanced AI models. As the AI landscape continues to evolve, it remains to be seen whether the high cost of o1-pro will be a barrier to its widespread adoption or if developers will be willing to pay the premium for improved performance.

In the broader context, the launch of o1-pro highlights the ongoing competition in the AI market, where companies are continually pushing the boundaries of what is possible with AI. As AI technology advances, it will be important to monitor the implications of these developments on the industry and society as a whole.