AI FOR SCIENCE · ON THE RECORD
Latent Labs
Builds generative AI models that design proteins such as antibodies from scratch, and tests the designs in its own wet lab in San Francisco.
Full funding record →On the timeline →People filter →
Research on the record
- 2026-07-15Latent Labs opens Latent-Y, an agent that designs antibodies from a written prompt, to any approved researcherLevel 4 · Tested in the physical world
- 2025-12-23Latent Labs designs antibodies from scratch that bind 9 of 18 targets; four designed against one of them raise no detectable immune response in ten donors' cellsLevel 4 · Tested in the physical world
- 2025-07-25Latent Labs' binder model hits 91-100% lab hit rates on three macrocycle targets, from 11 to 17 molecules assayed eachLevel 4 · Tested in the physical world
Graded by how far the evidence went, not how good the result is. See the methodology.
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.
Named on AI4S.fyi
Simon KohlFounder Latent Labs opens Latent-Y, an agent that designs antibodies from a written prompt, to any approved researcher
Sebastian SchmonMember of Technical Staff Latent Labs opens Latent-Y, an agent that designs antibodies from a written prompt, to any approved researcher- HHenry KenlayBuilds frontier models, in the team page's words Latent Labs designs antibodies from scratch that bind 9 of 18 targets; four designed against one of them raise no detectable immune response in ten donors' cells
Alex BridglandFounding Member of Technical Staff Latent Labs' binder model hits 91-100% lab hit rates on three macrocycle targets, from 11 to 17 molecules assayed each
Funding rounds tracked
- 2025-02-13Series A · $40MconfirmedLed by Radical Ventures, Sofinnova Partners · With Flying Fish, Isomer, 8VC, Kindred Capital, Pillar VC, Jeff Dean (angel), Aidan Gomez, Mati StaniszewskiStated use: its generative protein design models and partnerships
- 2025-02-13date disclosed · close date not givenSeed · $10MconfirmedLed by Pillar VC, Kindred Capital
Each round links to its line on the funding record, with its source.
Also in Protein design 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.
- 2026-10-28BioGeometrySeries A+ · not disclosed; the A series to date tops $100MBuilds generative models for biology: GeoFlow, an all-atom model that predicts the structures of proteins and their complexes and designs binders, and GeoCell, a virtual-cell model of how cells' gene expression responds to genetic and drug perturbations. An automated lab expresses and tests the designs and feeds the data back. It designs antibodies, vaccines, peptides and enzymes for partners and its own pipeline, including synthetic-biology products.
- 2026-09-24BigHat BiosciencesSeries C · $75MDesigns antibodies with machine learning tied to a fast automated wet lab (Milliner) that makes and tests designs; its lead drug BHB810 is in the clinic.
- 2026-08-11Aureka BiotechnologiesSeries B · $100MDesigns antibodies with its own biological foundation model, AuraIDE (released in open form as OpenDDE), trained and retrained on data from its own single-cell functional screens.
- 2026-07-14Chai DiscoverySeries C · $400MBuilds AI models that predict how molecules bind and design antibodies from scratch, and licenses them to drugmakers as software.
- 2026-06-17MoleculeMindSeries A (extension) · A-series cumulative total over US$100 millionAI protein structure prediction and design (MoleculeOS, MMFold, MMDesign) sold to drugmakers and industrial-enzyme users, with a few in-house early programs.
- 2026-03-20Helixon (Earendil Labs)Round (unnamed) · $787MAI protein modelling plus high-throughput experiments to design antibodies, especially bispecifics, which it licenses to Sanofi and Genentech and develops in-house.
This page compiles what AI4S.fyi has recorded and links each item to its source. It is not a grade of the organization and not its complete history: a blank means the record has not caught up as often as it means nothing happened.