In response to pressure from rivals including Chinese AI company DeepSeek, OpenAI is changing the way its new AI model, O3-Mini, communicates its “thought” process step by step.
On Thursday, OpenAI announcing OpenAI’s free and paid AI-powered chatbot platform, Chatbot in AI Testing, will display an updated “chain of thought” that shows more of the model’s “reasoning” steps and how it arrived at answers to questions. Subscribers to premium ChatGPT plans who use O3-Mini in the “High Reasoning” setting will also see this updated readout, according to OpenAI.
“We’re introducing an update to O3-Mini’s chain of thought designed to make it easier for people to understand how the model reasons,” an OpenAI spokesperson said. “With this update, you can follow the model’s reasoning, giving you more clarity and confidence in your answers.”
Image: OpenAI
Reasoning models like O3-Mini thoroughly verify themselves before giving results, which helps them avoid some of the pitfalls that models typically encounter. On the other hand, reasoning models take a bit longer to arrive at answers, typically seconds or minutes.
DeepSeek’s R1 model, an online “reasoning” model from O3-Mini, reveals its entire thought process, which many AI researchers argue is the preferred approach. In addition to making the model easier to study, the reasoning steps offer a better user experience in certain situations, helping to indicate when the model might be on the right or wrong track.
OpenAI had chosen not to show the full reasoning steps for O3-Mini and its predecessors, O1 and O1-Mini, in part due to competitive reasons. Instead, users only saw summaries of the reasoning steps — summaries that were sometimes wrong.
When we told people about the pre-O1-Preview release, seeing the live source was usually the “AHA” moment for them that made it clear that this was going to be a big deal. These aren’t the raw sources, but it’s a big step forward and I’m glad we can share that experience with the world. https://t.co/72zpprhmfk
– Noam Brown (@Polynoamial) February 6th 2025
OpenAI isn’t showing O3-Mini’s full reasoning steps yet, but the company said it “found a balance”: O3-Mini can “think freely” and then organize its “thoughts” into more detailed summaries.
“To improve clarity and safety, we’ve added an additional post-processing step where the model reviews the raw chain of thought, removing any unsafe content and then simplifies any complex ideas,” the OpenAI spokesperson continued. “Additionally, this post-processing step allows non-English speaking users to receive the chain of thought in their native language, creating a more accessible and friendly experience.”
In a Reddit AMA last week, OpenAi's chief product officer Kevin Weil hinted that change was coming.
“We are working on showing a lot more than we do today by showing the thought processes. It will be very, very soon,” he said. “Showing the whole chain of thought leads to competitive distillation, but we also know that people (at least power users) want that, so we will find the right way to balance it.”
and select 