Isomorphic Labs

What it is

Isomorphic Labs is a company applying artificial intelligence to drug discovery. It was launched out of DeepMind, and many of its team contributed to the creation of AlphaFold^[1]. Demis Hassabis is founder and chief executive, and Max Jaderberg is president^[1]. The company dates its own start to 2021, when AI drug discovery was still a hypothesis rather than a working process^[2].

That timing matters for how the company describes itself. Rather than taking an existing discovery pipeline and speeding it up with software, it set out to build a computational route into medicine design from the ground up, at a point when the idea had not yet been demonstrated in practice^[2].

The Drug Design Engine

Its platform is called the Isomorphic Labs Drug Design Engine, or IsoDDE^[2]. The company describes the engine as target- and disease-agnostic: it is built to work across therapeutic areas and modalities rather than being tailored around one class of biological target^[1]. In practice that means the same underlying system is meant to be pointed at different kinds of disease and different kinds of drug, instead of being rebuilt for each programme.

The engine sits downstream of structure prediction. The company says it developed and released AlphaFold 3 in May 2024, and then built further models to cover the parts of drug design that structure prediction does not reach^[1]. Knowing the shape of a protein target is one thing; deciding which molecules to make, and which of them are likely to work as drugs, is a wider problem, and the further models are aimed at that wider problem.

Running underneath is what the company calls its "dataverse", a curated life-science dataset used to run large volumes of in-silico experiments across GPU fleets^[1]. In-silico experiments are experiments carried out in simulation on computers rather than at the laboratory bench, and doing them at volume is what lets a model be tested against many candidate designs before any of them are made.

From small molecules to biologics

The company's programmes have expanded from small molecules to biologics, alongside an internal pipeline focused on oncology and immunology^[1]. Small molecules are typically compact synthetic chemicals, while biologics are larger molecules produced by living cells, such as antibodies; the two are designed, tested and manufactured in quite different ways. Covering both is consistent with the cross-modality ambition the company states for the engine^[1].

Partnerships

Isomorphic Labs has announced strategic research collaborations with several large pharmaceutical companies. It announced a collaboration with Novartis in January 2024 and expanded it in February 2025 by adding up to three more programmes^[3]. It has a strategic collaboration with Eli Lilly focused on small-molecule therapeutics against undisclosed targets^[3]. It also has a multi-target, cross-modality collaboration with Johnson & Johnson covering small molecules and biologics^[3].

These deals follow a common pattern in the field: the AI company supplies the computational platform and drug-design capability, while the pharmaceutical partner brings disease knowledge, clinical development and commercial reach.

Financing

On 12 May 2026 the company announced a $2.1 billion Series B round led by Thrive Capital^[4].

Where it sits

Isomorphic Labs is one of the clearest attempts to turn the protein-structure work that came out of DeepMind into an industrial drug-design process^[1]. Its own account of its history sets out the distance travelled: from an idea treated as a hypothesis in 2021 to a platform with named partners, a stated pipeline in oncology and immunology, and programmes spanning small molecules and biologics^[2][1][3]. Whether that route produces medicines is the question the company was set up to answer, and the collaborations and financing are the means by which it is being pursued.