OpenAI released its official news today, announcing the launch of its latest high-performance inference model, O1-Pro. OpenAI said the model has significantly improved performance and performance and is currently open to some selected developers.
According to OpenAI, the O1-Pro model is available for specific developers located in layers 1 to 5, and has powerful features, including support for visual input, function calls and structured output, while collaborating seamlessly with OpenAI's response and batch API.
However, performance improvements are accompanied by higher cost of use. OpenAI announced the pricing of O1-Pro, with the price of its input model at $150 per million tokens, while the price of model-generated content at $600 per million tokens. Compared with other models previously released by OpenAI, O1-Pro is priced significantly higher.
Despite its high price, the enhanced performance and rich features offered by O1-Pro, such as visual understanding and more flexible API interaction, are expected to attract developers who need to handle complex tasks and have higher demands on model performance. The launch of O1-Pro also further demonstrates OpenAI's efforts to continue to promote the cutting-edge of artificial intelligence technology and strategies to provide diversified model selection based on different developers' needs. At present, OpenAI has not announced the timetable for the full opening of O1-Pro, and interested developers need to pay attention to the official follow-up notice.