ChatMDR 4 AI models explained

/ Guides, News / By

ChatMDR 4 models explained

As of ChatMDR, ChatIVDR, ChatFDA and ChatAIAct version 4 (April 30, 2025) there are three different AI models available, each with their own characteristics. Here’s how to make the most of each.

First of all, the obvious: how to select each model. This is done using the model selector, which is at the left of your screen, or at the top on mobile:

Desktop user interface of ChatAIAct, with AI model selector highlighted.
Screenshot
Mobile user interface of ChatMDR, with AI model control highlighted
Lightbulb, identifying the Smart model

Smart

Smart is the workhorse model. As its name suggests, it’s pretty clever and can generate good answers for most of your questions. It’s great at conversation, so keep talking to it (so to speak) to get deeper insights.

Good for

  • General questions
  • Looking up information in regulation or guidances
  • Writing parts of documents

Less suitable for

  • Developing strategies or procedures
  • Checking detailed answers or rationales
Partial diamond, identifying the Diamond model

Diamond

Diamond is a reasoning model. That means that, internally, it uses some time to think before it answers. Its deliberations can be seen by clicking on a little Chain of thought link. The reasoning ability allows it to evaluate various alternatives before jumping to a conclusion. It is very good for structural work.

Good for

  • Coming up with strategies or rationales that take elements of the regulation
  • Checking answers and rationales
  • Outlining an approach to something, rather than giving the answer itself.

Less suitable for

  • Information lookup
  • Writing easy to read text

Pro tip: switch model mid-chat and let Diamond verify your Smart answers! See an example here.

Flying graduation hat, indicating the Scholar model

Scholar

Scholar is a hybrid of a reasoning model and a smart model, aiming to be the best of both worlds. It takes your user input and turns it into a more elaborate strategy for an expanded version of the Smart model to take a look at. The reason this works is that the second model will have a much better idea of which parts of the regulations to include and how to apply them.

Good for

  • All types of questions, strategies and rationales

Less suitable for

  • Quick results (because the initial answer of the model takes longer)

Plus and Unlimited Users: the use of the Scholar model is for paying customers only.