OpenAI announced that it will allow third-party developers to use itsThe latest artificial intelligence model o1In addition, more application tools are also launched for developers.
O1 is currently OpenAI's most powerful artificial intelligence model. During its calculation process, it considers the nature of the problem and the correctness of the answer. Therefore, it performs better in physics, chemistry, biology, mathematics, and programming. However, it takes more time to verify the execution speed. Unlike the GPT series of models, which mainly draw conclusions through data comparison, O1 will focus more on whether the answer is correct and whether it matches the logic of the question.
In answering the International Mathematical Olympiad test, OpenAI said that GPT-4o's answer accuracy was only 13%, but o1's accuracy increased to 83%.
This announcement allows all third-party developers to access the o1 model via an API, adding structured output, function calls, visual reasoning, and fine-tuning for rationality. OpenAI has also adjusted the price of using the Realtime API, for example, reducing the price of GPT-4o audio output by 60%, and adjusting the input price of GPT-4o mini to $100 per 10 million tokens and the output price to $20.
Other parts include adding preference fine-tuning functions to make the model output results more in line with expected needs. At the same time, it also provides new tool resources for Go and Java language developers to help more developers create artificial intelligence application services.
We're bringing OpenAI o1 to the API. We're rolling out access to developers on usage tier 5 starting today, and rollout will continue over the next few weeks.
o1 supports:
⚙️ Function calling
🗂️ Structured Outputs
👀 Vision
📝 Developer messages
🧠 Reasoning effort pic.twitter.com/Ax8TT0IRke— OpenAI Developers (@OpenAIDevs) December 17, 2024
