Radical Numerics

Research on the record

• 2026-10-20Radical Numerics reports that phage variants its Omnii-1 model designed were given to one patient with a drug-resistant infection, under compassionate useLevel 4 · Tested in the physical world
• 2026-10-20Radical Numerics reports its Omnii-1 model scoring above AlphaGenome and Borzoi at telling which changes in non-coding DNA actually do somethingLevel 3 · Checked against existing evidence
• 2026-10-20Radical Numerics and Aptamino run three rounds of lab-measured antibody design with Omnii-1, and two designs bind tighter than the test can readLevel 4 · Tested in the physical world
• 2026-10-20With Lawrence Livermore, Radical Numerics reports its biology model predicting which rare earth metal a protein prefers better than the other methods testedLevel 2 · Computed
• 2026-06-15Radical Numerics previews Omnii, a genome language model that scores noncoding variants above CADD and Evo 2 on the company's own benchmark runsLevel 2 · Computed

Graded by how far the evidence went, not how good the result is. See the methodology.

Also on the timeline

• 2026-06-15Radical Numerics, started by the team behind the Evo and Evo 2 DNA models, launches with a $50M seed round led by Emergence Capital to build one model that reads DNA, RNA and proteins togetherCapital

People

Only people the AI4S.fyi record has named: in a round, a timeline entry, commentary, an interview or a sourced role. Not the full team.

Funding rounds tracked

• 2026-06-15Seed · $50MconfirmedLed by Emergence Capital · With Obvious Ventures, Triatomic Capital, Factory, First Spark VenturesStated use: scale the next generation of models, grow the research team, and build compute and data infrastructure

Each round links to its line on the funding record, with its source.

Also in Science models and AI scientists on the funding board

The other companies the funding board files under the same field, latest round first. A list of what the record tracks, not a ranking or a market map.

• 2026-10-13Tianxu Intelligent TechnologyPre-A+ · tens of millions of yuanBuilds a weather model for the energy industry (wind speed and solar irradiance rather than general forecasts) with the China Meteorological Administration's public service centre, and is putting the model on satellites of its own to make the observations it trains on.
• 2026-10-13Drylab AISeed · $3MAgents that run computational biology workflows end to end (planning the analysis, choosing among more than 800 tools and databases, getting the compute and working through failures), with a separate model of its own checking each analytical decision and flagging method errors; installed on a customer's private infrastructure or workstations.
• 2026-10-06VinciSeries B · $250MBuilds a physics foundation model with GPU-native solvers that gives engineers solver-accurate thermal, thermo-mechanical and fluid results while a design is still changing, starting with semiconductors.
• 2026-09-30ZenithonSeed · $10MWorld models for extreme physics (fusion reactors, rockets and semiconductor manufacturing) meant to replace simulations that take days to months, starting with fusion customers.
• 2026-09-30Autopoiesis SciencesSeed · $11MBuilds AI for scientific reasoning: Aristotle, a research assistant that helps scientists form hypotheses and weigh evidence, and Columbus-1, an autonomous system for long research problems. It says it builds training environments and datasets from expert scientists' judgment.
• 2026-09-24Metis ReasoningRound (unnamed) · tens of millions of dollarsEarly-stage lab building AI meant to reason and act reliably beyond its training data, naming embodied AI, science and engineering as target fields.

All 25 in this field →