During a continuous live broadcast event lasting 12 working days, OpenAI released the API for the o1 model on the 9th day for developers, announcing a major upgrade to the real-time API that supports WebRTC technology.
From the release date, OpenAI will provide access to the o1 API for developers using API level 5. Compared to the previous preview version, the updated o1 model API has achieved a 60% reduction in reasoning costs and introduced advanced visual processing capabilities. Additionally, the cost of audio processing for GPT-4o has also decreased by 60%, while the price of the mini version has dropped significantly by 10 times.
The o1 model API also introduces a new API parameter called "reasoning_effort," which allows developers to control the amount of time the model spends thinking before answering questions. Furthermore, OpenAI has launched a brand new preference fine-tuning technique that enables large models to better adapt to users' personalized styles through direct preference optimization algorithms.
OpenAI's weekly active user count has surpassed 300 million, with users sending over 1 billion messages to ChatGPT daily. Compared to August of this year, when the weekly active user count had just crossed 200 million, this shows a rapid growth in OpenAI's user base. The release of these updates and data not only demonstrates OpenAI's technological advancements in the field of artificial intelligence but also reflects its widespread influence among global users.
