Credit: Pexels
OpenAI’s latest AI model promises better reasoning and reliability but comes with a steep price tag
San Francisco, CA – OpenAI has officially launched o1-pro, a souped-up version of its o1 “reasoning” AI model, now available in its developer API. This premium AI promises more reliable and accurate responses, but it comes with a hefty price tag—making it OpenAI’s most expensive model to date.
Smarter AI at a Premium Price
According to OpenAI, o1-pro uses more computing power than its predecessor, enabling it to “think harder” and provide consistently better answers to complex problems. The model is currently available only to select developers—specifically those who have already spent at least $5 on OpenAI API services.
But exclusivity isn’t the only thing that sets o1-pro apart—it’s also the cost. OpenAI is charging:
- $150 per million input tokens (roughly 750,000 words)
- $600 per million output tokens
That’s twice the price of OpenAI’s GPT-4.5 for input and a whopping 10 times the price of regular o1. Clearly, OpenAI is banking on o1-pro’s improved performance being worth the steep investment.
Why Pay More for o1-Pro?
So what makes o1-pro special? OpenAI claims it provides more reliable responses to the toughest problems, which could be a game-changer for developers working on AI-driven applications requiring high accuracy.
“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 told TechCrunch. “After getting many requests from our developer community, we’re excited to bring it to the API to offer even more reliable responses.”
Early Reviews Are Mixed
While OpenAI is confident in o1-pro’s capabilities, early user experiences have been mixed. The model has been available in ChatGPT for Pro subscribers since December, but some users found it struggled with Sudoku puzzles and even optical illusion jokes.
Additionally, OpenAI’s own internal testing from late last year showed only slight improvements over the standard o1 in coding and math problems—though the company claims it answers those questions with greater consistency.
Is o1-Pro Worth It?
For developers looking for top-tier AI performance, o1-pro might be worth the premium. But for those on a budget, the question remains: are the improvements significant enough to justify the high cost?
As more developers gain access to o1-pro, we’ll soon see if OpenAI’s most expensive AI model truly lives up to the hype.
