Best AI Drug Discovery Companies 2026

Isomorphic Labs represents the most capital-efficient path from Nobel Prize-winning biology to drug discovery. Spun out of Google DeepMind in 2021 under CEO Demis Hassabis, the company is built on AlphaFold 3 — the system that predicted the three-dimensional structure of virtually every known protein and earned its creators the 2024 Nobel Prize in Chemistry. AlphaFold 3 extends the capability to model interactions between proteins, DNA, RNA, and small molecule drugs, providing the structural foundation for designing molecules that will bind their targets with high precision.

The company's proprietary generative design models — built on top of AlphaFold's structural knowledge — propose candidate therapeutics for disease targets including those previously considered undruggable due to their structure. In January 2025, Isomorphic announced landmark partnerships with Eli Lilly ($45 million upfront, up to $1.7 billion in milestones) and Novartis ($37.5 million upfront, up to $1.2 billion in milestones) — both structured around Isomorphic's AI designing the candidate molecules. The total deal value of nearly $3 billion is the largest AI drug discovery pharma partnership in history.

In May 2026 the company raised $2.1 billion in a Series B led by Thrive Capital — joined by Alphabet, GV, Temasek, CapitalG, MGX, and the UK Sovereign AI Fund — to fund platform scaling and pipeline advancement toward IND filings. First clinical trials are expected in late 2026. For pharma companies evaluating AI partners, Isomorphic offers the most credible combination of scientific pedigree (AlphaFold), capital scale ($2.7B+ raised), and demonstrated willingness to structure large milestone-based collaborations.

Recursion Pharmaceuticals is the most clinically advanced AI-native drug discovery company in the world by number of programs in human trials. Founded in 2013 at the University of Utah by Chris Gibson and Blake Borgeson, Recursion's OS (operating system) industrialises drug discovery by mapping biology at unprecedented scale — the company has generated over 50 billion cellular measurements — and using this biological map to identify drug-like relationships that would be invisible to human researchers.

In November 2024, Recursion completed the $688 million all-stock acquisition of Exscientia — Oxford's leading AI drug design company, which pioneered AI-designed molecules in human trials. The merged entity combines Recursion's 60+ petabytes of multimodal biological data (cellular imaging at scale, genetic perturbations, metabolomics) with Exscientia's generative chemistry capabilities (AI-designed small molecules with predicted binding affinity), creating an end-to-end AI drug discovery platform. The combined pipeline includes REC-3964 (Phase 2, C. difficile infections), REC-1245 (Phase 1, oncology), and REC-3565 (Phase 1), with over 10 partnered programs generating $450M+ in milestones received and $20B+ in potential future milestones.

Strategic partnerships with NVIDIA (compute infrastructure), Roche, Bayer, and others validate the platform's commercial appeal. For pharma companies evaluating AI partnerships, Recursion offers the deepest clinical track record, the largest proprietary biological dataset, and the broadest coverage of the drug discovery stack through the Exscientia merger.

Insilico Medicine owns the most important milestone in AI drug discovery history: it is the first company to advance a molecule where both the therapeutic target and the drug candidate were identified entirely by generative AI through Phase 2a clinical trials, achieving statistically meaningful efficacy. Founded in 2014 by Alex Zhavoronkov, Insilico operates a trilogy of integrated AI systems: PandaOmics (target identification from multi-omics data), Chemistry42 (de novo molecule generation), and inClinico (clinical outcome prediction).

The lead asset, rentosertib (ISM001-055), is a TNIK inhibitor for idiopathic pulmonary fibrosis — a progressive and universally fatal lung-scarring disease with no approved disease-modifying therapies. In the Phase IIa trial (71 patients, 12 weeks), the 60 mg dose achieved a mean improvement in forced vital capacity of +98.4 mL, while the placebo group declined by −20.3 mL — a difference of approximately 118 mL that is considered clinically meaningful in IPF. The results were published in Nature Medicine, independently validated, and presented at the American Thoracic Society Annual Meeting 2025.

In December 2025 Insilico became the first AI-native drug discovery firm to conduct a major exchange IPO, raising $293 million on the Hong Kong Stock Exchange at a market cap of approximately $2.7 billion. A rentosertib inhalation solution has received IND clearance for direct-to-lung delivery, expanding the therapeutic application. The company's Hong Kong IPO and Phase 2a results together represent proof-of-concept for the entire AI drug discovery sector.

Xaira Therapeutics is the most boldly capitalised new AI drug discovery entrant — launched in 2024 with $1 billion in Series A funding from ARCH Venture Partners and Foresite Labs at a $4 billion valuation, the largest single venture raise for an AI biotech startup. The company was co-founded by David Baker, the 2024 Nobel Prize in Chemistry laureate whose work on protein design and directed evolution underlies the entire modern AI structural biology field, alongside CEO Marc Tessier-Lavigne, former President of Stanford University.

Xaira's defining architectural choice is AI-first: the company built its AI platform before any drug pipeline, meaning every future therapeutic candidate will be generated by the models rather than the models being trained on existing company chemistry. The flagship AI model, X-Cell, is described as the largest cell perturbation model ever built — trained to predict gene function from large-scale genetic perturbation experiments, enabling target identification from the biological consequences of gene silencing across thousands of cellular contexts. This approach is designed to identify truly novel disease targets that conventional hypothesis-driven biology would not prioritise.

Xaira is the longest-horizon bet among the companies in this guide — it will be 2026–2028 before its AI-generated pipeline reaches IND-enabling studies — but it has the most scientifically credentialed founding team and the most radical commitment to building foundation models for biology before building drugs. For pharma companies and investors evaluating early-stage partnerships, Xaira represents the clearest expression of the thesis that AI foundation models for biology will be as transformative as LLMs for language.

insitro brings the most credentialed AI leadership in the drug discovery sector. CEO Daphne Koller — Stanford Professor, Coursera co-founder, and recipient of the MacArthur 'Genius Grant' — founded the company in 2018 on the thesis that the same machine learning methods transforming language and vision could systematically accelerate drug discovery if applied to sufficiently rich biological datasets. The company has raised approximately $800 million from a16z, Google Ventures, Third Rock Ventures, and others.

In January 2026, insitro launched TherML (Therapeutic Machine Learning) following the acquisition of CombinAbleAI, creating a unified AI platform covering all four major drug modalities: small molecules, oligonucleotides, antibodies, and complex biologics. This modality-agnostic architecture is unique — most AI drug discovery companies focus on small molecules only — and positions insitro to pursue the broadest range of biological targets. The approach begins with generating high-resolution multimodal biological data (genomics, cellular imaging, transcriptomics, phenotypic screens) and applying ML to identify disease mechanisms and design therapeutics predicted to work in the relevant human cellular context.

Pharma partnerships validate the platform: Eli Lilly expanded its collaboration in September 2025 to cover ML models predicting pharmacological properties of small molecules; Bristol Myers Squibb collaborates on ALS target identification. The company's programmes focus on metabolic disease and neuroscience — two areas where biological complexity has historically defeated conventional drug discovery. For pharma companies seeking a partner with academic ML rigour, multi-modality coverage, and demonstrated ability to deliver computational assets to Big Pharma, insitro is the strongest option.

Schrödinger is the most commercially mature and revenue-generating company in the AI drug discovery landscape. Founded in 1990 as a computational chemistry software company and taken public on NASDAQ in 2020, Schrödinger has built the deepest pharma customer base in the sector: every single one of the world's top 20 pharmaceutical companies by revenue licences the Schrödinger platform, generating $201 million in annual contract value. Q1 2026 drug discovery revenue rose 124% year-over-year to $22.9 million as AI-assisted drug design reached mainstream pharmaceutical adoption.

Schrödinger's technical differentiation is the integration of physics with machine learning. The company's free energy perturbation (FEP+) technology uses quantum mechanical simulations to compute binding affinities with high accuracy — a level of ground-truth reliability that pure ML models cannot achieve without experimental data. This physics foundation is complemented by ML models that accelerate the screening of chemical space and predict ADMET properties. In 2026 the company unveiled Bunsen, an agentic AI co-scientist that autonomously executes complex molecular discovery workflows — enabling researchers to direct multi-step computational campaigns in natural language and reducing the need for deep computational chemistry expertise.

A new Eli Lilly partnership for AI drug discovery joins existing collaborations with Bristol Myers Squibb, Takeda, and others. For drug discovery teams evaluating computational platforms, Schrödinger is the lowest-risk choice: it has the most customers, the deepest validation across pharma workflows, and the clearest path from software licensing to platform-generated drug value through its equity-sharing model with biotech collaborators.