January 2026 Release Notes

🧩 Structured Data: Bring it all together

Many abstraction workflows rely on a mix of structured data and abstracted data. Now, Brim supports that end-to-end.

You can upload structured data like demographics, lab values, or other tabular inputs and then:

  • Use structured data as input to abstracted variables
  • Use structured data values to filter during generation and review
  • Export structured and abstracted data together


This makes it easier to ground abstractions in known facts, reduce unnecessary generation, and produce cleaner, more useful outputs.

You can get started by going to Project Setup > Project Data > Add Data. Read more in our documentation.

🤖 Configure LLMs by Project

You no longer have to settle for one LLM and one Advanced LLM across your entire workspace.

Brim now lets you configure multiple LLMs in Settings > Admin > LLM Config, and assign specific models to specific projects. This makes it easy to:

  • Customize compute usage by project
  • Experiment with different models on the same data
  • Balance cost, speed, and quality depending on the task

Read more in our documentation.

🎨 Design & Usability Improvements

We’ve made a number of visual and interaction improvements throughout Brim to make the tool easier and more pleasant to use, including:

  • Improved contrast for better readability
  • More consistent interaction patterns
  • Helpful icons and visual cues

No new concepts to learn; just a smoother experience.


Also in this Release:

  • Duplicate Project from the Admin > Projects page.
  • Scope icons throughout the experience
  • Unique names enforced across variables, dependent variables and structured data variables
  • Improved reasoning for aggregated variables
  • Fix for cancel invite link failing for projects with 0s
  • Performance Improvements on Single Note API.

✨ How to Get the Latest Brim Version

This version is called: 2026.01.28

You can check your current version in Settings > About.

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.