OpenAI rolled out interactive math and science modules to ChatGPT on March 10, 2026. When you ask about a supported concept, ChatGPT now surfaces a live module alongside the text explanation, letting you adjust variables and watch the outputs change in real time. This is not a separate app or a standalone tutor product. It is a visual layer built into the same chat interface 140 million people already use each week for math and science help.
What the Interactive Modules Actually Do
The mechanic is direct. You ask about a supported topic and instead of only getting a text explanation, you get a module where you can manipulate inputs and see results update live. Adjust the sides of a triangle in the Pythagorean theorem and the hypotenuse updates. Change the pressure or temperature in the ideal gas law and the other variables respond. The point is to let you build intuition about how variables relate to each other rather than just memorizing that a formula exists.
OpenAI launched with over 70 topics covering high school and college-level material. The list includes Hooke’s law, kinetic energy, Coulomb’s law, Ohm’s law, compound interest, exponential decay, the lens equation, slope-intercept form, trig angle sum identities, Charles’ law, the period-frequency relation, cone and cylinder volume formulas, and more. It covers a solid portion of what students actually get stuck on.
Early feedback from testing was that students got a better grasp of how variables relate. Educators noted a shift away from formula memorization toward understanding why a formula behaves the way it does. That tracks. Watching a graph change as you move a slider is a different cognitive experience than reading that y increases as x increases.
Where This Fits in ChatGPT’s Learning Arc
This builds on the study mode OpenAI shipped in July 2025, which added step-by-step guidance, Socratic-style prompts, quizzes, and skill-level calibration. That was about pacing and scaffolding. This new layer is about visual experimentation. OpenAI flagged the visual component as a next step when study mode launched, so this is a planned continuation rather than a surprise pivot.
It also connects to OpenAI’s broader education push, which includes NextGenAI and their Learning Lab work measuring how AI affects learning outcomes over time. The company has been deliberate about moving into the education market and this is one of the more concrete product moves in that direction.
For context on where ChatGPT fits in the broader AI tool ecosystem, ChatGPT Pro vs Claude Max covers how the subscription tiers compare if you are evaluating which plan makes sense for your use case.
Topic Coverage at Launch
The 70+ topics span four main subject areas. Physics and math make up the bulk of the list, with chemistry and applied math topics like compound interest rounding it out. Here is an approximate breakdown of how the launch topics distribute across those areas:
Approximate distribution of the 70+ supported topics by subject area at launch
Some specific examples from each area give a sense of the depth. On the math side you get linear equations, slope-intercept form, triangle area, circle area, circumference and equation, the binomial square, difference of squares, and surface area of a sphere. Physics covers kinetic energy, potential energy, Hooke’s law, Ohm’s law, Coulomb’s law, the lens equation, and the period-frequency relation. Chemistry includes the ideal gas law and Charles’ law. Applied math brings in compound interest and degrees of freedom.
Availability
The rollout covers all logged-in ChatGPT users globally on both consumer and business plans starting March 10, 2026. No paywall, no waitlist. OpenAI says they plan to expand to more subjects over time, though no specific topics or timeline were given for what comes next.
If you use ChatGPT regularly for any STEM work, the feature surfaces automatically when you ask about a supported topic. You do not need to enable anything. Just ask about the Pythagorean theorem or kinetic energy and the module will appear if your account has the rollout.
This is a useful addition for students, parents helping with homework, and anyone who learns better by manipulating variables than by reading a static explanation. It does not change what ChatGPT is at its core. It adds a better format for a category of questions the model already handled with text alone, and for that specific use case, it is a clear improvement.