With the launch of o3-pro, let’s talk about what AI “reasoning” actually does
1 min read
Summary
OpenAI has made a more capable version of its ChatGPT Pro and Team users, called o3-pro, and also cut the price for using the model by 87 per cent.
The new model includes capabilities such as web search, file and image analysis, and Python execution, aimed at complex analytical tasks.
However, it’s slower than the previous o1-pro model and OpenAI recommends it be used where accuracy is more important than speed.
The company has also reduced costs across the board, lowering the price of the standard o3 model by 80 per cent and meaning o3-pro is now 87 per cent cheaper to use than o1-pro.
This makes the higher capability reasoning models more affordable for developers and users.
