About

We came together through a series of unconventional technical bets, from long-context architectures with Hyena to modeling and generating DNA with Evo and Evo 2. When existing approaches fell short, we built new architectures and systems. We started Radical Numerics to take that work further.

We’re building toward general biological intelligence. Our model, Omnii, learns across DNA, RNA, proteins, and other biological signals to read, write, and interpret biology: predict biological effects, generate sequences, and identify relationships that guide research.

Biology is a whole-system problem. Our thesis is that learning these signals together provides the context needed for better predictions, stronger candidate designs, and insights unavailable from any one signal alone.

Our mission is to solve biology. We want to help people prevent and treat disease, create new biological functions, and expand what can be engineered with biology.

We measure progress through scientific evaluation and experimental results. Ultimately, success means lives saved, diseases cured, and new biological capabilities made possible.

We built Radical Numerics to advance biological intelligence and strengthen biosecurity together. The capabilities that enable new medicines and biological designs can also be misused. Our research pursues both the ability to engineer biology and the ability to identify and respond to biological threats.