Radical Numerics
Radical Numerics is an AI research lab building general biological intelligence through multimodal models trained on biological data.
- 16
- San Francisco, CA, USUnited States
Radical Numerics is an AI research lab building general biological intelligence through multimodal models trained on biological data.
Which advantages over its rivals hold up, and why are they hard to copy?
The founding team previously built Evo and Evo 2, the first AI models capable of generating DNA sequences at scale and trained on the genomes of more than 100,000 species. This track record provides strong scientific credibility in generative genomics.
Radical Numerics has assembled a scientific advisory board that includes Eric Horvitz of Microsoft, Chris Re of Stanford, George Church of Harvard, and Andrew Weber, formerly of the United States Department of Defense. These advisors span machine learning systems, genomics, and biodefense.
Is its category moving in its favor or against it?
AI applied to drug discovery is projected by Fortune to reach 25 billion United States dollars by 2035, with biological foundation models attracting growing pharmaceutical and biotech investment. Demand for multimodal models that read and write biology continues to expand alongside larger biological datasets.
Biodefense represents an adjacent market driven by national security interest in early pathogen detection. Radical Numerics is partnering with a national laboratory on detecting and characterizing pathogens, including AI-generated ones.
What does the product do for its buyer, and how is it delivered?
Radical Numerics develops multimodal AI models that learn directly from biological data, including DNA, RNA, and proteins. The company is scaling context windows from one million tokens toward one billion tokens for biological sequences.
Its initial research surface is Omnii, a genomic language model intended to support drug discovery, pathogen detection, and biological design. The team plans to monetize through API licensing, fine-tuned proprietary models for pharmaceutical partners, and milestone payments.