Radical Numerics Launches with $50M Seed Round to Build General Biological Intelligence for Drug Discovery and Biodefense
核心洞察
Radical Numerics (搜索), an AI research lab founded by the creators of generative genomics, launched with $50 million in seed funding led by Emergence Capital (搜索).
The company is building multimodal AI models that learn across DNA, RNA, and proteins simultaneously, aiming to unify biology into a single general biological intelligence.
Early results for its next-generation model Omnii (搜索) show state-of-the-art performance in identifying causal regulatory variants linked to Alzheimer's disease (搜索) and detecting AI-generated pathogens.
Radical Numerics (搜索), an AI research lab founded by the team that pioneered generative genomics, announced its official launch with $50 million in seed funding. The round was led by Emergence Capital (搜索), with participation from Obvious Ventures, Triatomic Capital, Factory, and First Spark Ventures. Pre-seed investors included Patrick Collison, CEO of Stripe and cofounder of the Arc Institute. The San Francisco-based company is building what it calls "general biological intelligence"—a new class of AI that learns directly from biological data across DNA, RNA, proteins, and beyond, unifying all dimensions of biology into a single multimodal model.
The founding team includes CEO Eric Nguyen (PhD, Stanford Bioengineering & AI), Chief AI Scientist Michael Poli (PhD, Stanford; Liquid AI founding team), President Stefano Massaroli (postdoc with Yoshua Bengio; Liquid AI founding team), and CTO Armin Thomas (postdoc, Stanford with Chris Ré; Liquid AI). Scientific advisors include Eric Horvitz, Chief Scientific Officer at Microsoft; Chris Ré at Stanford; George Church at Harvard; and Andrew Weber, former U.S. Assistant Secretary of Defense for Nuclear, Chemical, and Biological Defense Programs.
From Generative Genomics to a Company
The founders created the field of generative genomics with Evo (搜索) and Evo 2, the first AI models capable of both reading and writing DNA at scale, trained on the genomes of more than 100,000 species. Evo and Evo 2 were featured on the cover of Science magazine, in Nature, and at TED2025. The model was used to design novel CRISPR systems, and external scientists later used it to generate the first complete AI-designed genome: a bacteriophage, a virus that infects bacteria and is not harmful to humans.
"It still wasn't being picked up in the way we thought it would," Nguyen told Fortune of the academic work. "So we basically said: we have to show the recipe."
That milestone—the world's first fully AI-designed functional virus—pushed the team to build a company. "Evo (搜索) showed that AI can generate DNA and whole genomes, the next generation of models will go further with the ability to control function, and eventually, create entirely new forms of life," Nguyen said.
Omnii (搜索): Next-Generation Genomic Language Model
Alongside its launch, Radical Numerics (搜索) is previewing Omnii (搜索), its next-generation genomic language model. Early results show Omnii setting a new state of the art in identifying causal regulatory variants and transferring zero-shot to experimental settings. Without specific training, Omnii recovers experimentally validated functional variants at loci associated with Alzheimer's disease (搜索). The same model also achieves state-of-the-art performance for detecting AI-generated or AI-manipulated pathogens.
The company's approach differs from most AI biology companies, which are single-modality—such as Isomorphic Labs focusing on proteins or Inceptive targeting RNA. Radical Numerics (搜索) is betting that the bottleneck in drug development lies in understanding how drugs behave inside an entire biological system. "Getting the drug made won't be the bottleneck forever," Nguyen said. "You have to understand the whole system."
Dual Mandate: Health and Biodefense
Radical Numerics (搜索) holds a dual mandate: to advance biological design for human health and build the biodefenses to protect it. The company is partnering with a cancer diagnostics company to apply its multimodal model to pancreatic and multi-cancer detection, combining multiple molecular signals into a single diagnostic that existing tools may miss. On the biosecurity front, Radical Numerics is partnering with a national laboratory to use its models to detect and characterize pathogens, whether naturally occurring or AI-generated.
"The same models that can help cure disease may also lower the barrier to designing harmful biology. These forces are inseparable," Nguyen said. "Biology will be the most consequential application of AI."
Gordon Ritter, Founder and General Partner at Emergence Capital (搜索), emphasized the company's integrated approach to safety: "Most labs bolt safety on at the end. Radical Numerics (搜索) built it into the foundation. They've paired frontier-model capability with real biosecurity expertise to open a scientific field that didn't exist before."
The company's revenue model is still taking shape but is described as a mix of API licensing, fine-tuned proprietary models for pharmaceutical partners, and milestone payments. "No one has figured out the right business model for how AI companies commercialize in life sciences," Nguyen said. "If anybody says they have a formula, they're just full of it."
With 98 percent of the human genome still not understood, Nguyen is betting that the same technology that could one day explain it could also protect against those who might exploit it. The new funding will be used for scaling the next generation of its models and expanding the team with frontier AI research talent.