OpenAI introduces o1-pro, a more powerful version of the popular o1 model.
OpenAI has launched the o1-pro model, which should provide “consistently better responses” than the original o1 model. This is because it uses more computing power and because reasoning models take the time to think about their response to the user’s prompt.
High price
For now, the model is only available to developers who have paid at least five dollars for OpenAI’s API services. The price is set at 150 dollars per million tokens. This equates to 750,000 input words and 600 dollars per million generated tokens. This makes o1-pro the most expensive model to date.
OpenAI doubles the price of GPT-4.5 (75 dollars per million tokens) for o1-pro, and it is ten times more expensive than the regular o1 model (15 dollars per million tokens). The only thing you get for that price is better computing power and thus better answers. The specifications are similar to those of o1.
Developers already excited
OpenAI thinks that developers are willing to pay that higher price. “O1-pro in the API is a version of o1 that uses more computing power to think harder and provide even better answers to the most difficult problems,” a spokesperson for the company tells TechCrunch. “After receiving many requests from our developer community, we are excited to bring it to the API to provide even more reliable answers.”
Whether regular users will be as enthusiastic remains to be seen. In December, o1-pro was already offered to ChatGPT Pro subscribers, but the model was not particularly well received. According to users, it was still far from perfect.