More than 200 AI-designed drugs are now in clinical development. Phase I success rates for AI-discovered compounds are running between 80% and 90%, roughly double the historical average. The first fully AI-designed drug has posted positive Phase IIa efficacy data in humans. And in the first three weeks of 2026 alone, Eli Lilly, GSK, and Pfizer each signed major AI platform deals.

The hype cycle is over. The clinical validation cycle has begun.

But here is the problem: there are hundreds of companies claiming to do AI drug discovery. Most of them are pre-revenue platform companies with no clinical assets. Some are rebranding basic computational chemistry as “AI” to attract funding. And the venture market has flooded the space with capital that has made it nearly impossible to distinguish signal from noise.

This guide cuts through it.

We profiled 25 AI drug discovery companies based on a single filter: clinical evidence. Every company on this list has either advanced AI-designed or AI-discovered molecules into human clinical trials, delivered clinical candidates to pharmaceutical partners, or demonstrated platform validation through major industry partnerships backed by significant capital. No vapor. No pitch decks. Just pipeline.

This guide is updated quarterly. Companies are evaluated based on clinical pipeline status, partnership depth, platform differentiation, and disclosed results. Suggest a company for review.

Why This List Matters Right Now

The AI drug discovery market was valued at approximately $1.9 billion in 2025 and is projected to reach $2.6 billion in 2026, growing at a 27% compound annual rate toward an estimated $16.5 billion by 2034. More important than the market size is what is happening clinically. An estimated 15 to 20 AI-originated drugs are expected to enter pivotal trials during 2026. Multiple analysts project a 60% probability that the first AI-designed drug receives regulatory approval by 2027.

The year 2026 is the defining test. Phase III results will determine whether AI can deliver drugs that work at scale, not just drugs that reach the clinic faster. Every company on this list is part of that test.

Tier 1: Clinical-Stage Validation

These companies have AI-designed or AI-discovered drugs in human clinical trials with disclosed data.

1. Insilico Medicine

Headquarters: Hong Kong / New York | Founded: 2014 Platform: Pharma.AI (generative AI for target discovery + molecular design) Most Advanced Asset: Rentosertib (ISM001-055) — Phase IIa completed

Insilico Medicine holds the most significant clinical proof point in the entire AI drug discovery field. Rentosertib is a first-in-class TNIK inhibitor for idiopathic pulmonary fibrosis where both the target and the molecule were discovered using generative AI. In the Phase IIa trial published in Nature Medicine, the 60 mg dose showed a 98.4 mL improvement in forced vital capacity compared to a 20.3 mL decline in the placebo group over 12 weeks. That is the first time an AI-designed molecule has demonstrated both safety and efficacy in a controlled human trial.

Beyond rentosertib, Insilico has nominated 22 preclinical candidates since 2021 across oncology, fibrosis, immunity, and age-related diseases. The company has signed software licensing deals with 13 of the world’s top 20 pharmaceutical companies, and in January 2026, announced a near-$120 million partnership with Qilu Pharmaceutical for cardiometabolic therapies. Insilico’s business model — combining internal pipeline, software licensing, and pharma partnerships — is arguably the most diversified in the sector.

What to watch: U.S. Phase IIa trial of rentosertib currently enrolling. Broader pipeline readouts across oncology and fibrosis expected through 2026.

2. Recursion Pharmaceuticals (+ Exscientia)

Headquarters: Salt Lake City, UT | Founded: 2013 Platform: Recursion OS (phenomics-driven discovery + precision chemistry) Most Advanced Assets: REC-3964 (Phase II), REC-1245 (Phase I), REC-3565 (Phase I)

The 2024 merger of Recursion and Exscientia created the most comprehensive AI drug discovery platform in the industry, integrating Recursion’s high-throughput biological imaging — capturing millions of cellular experiments per week — with Exscientia’s automated precision molecular design. Their supercomputer, BioHive-2, built in partnership with NVIDIA, is one of the most powerful computing systems in biopharma.

The combined pipeline includes REC-3964 for C. difficile infection (Phase II update expected Q1 2026), REC-1245 targeting solid tumors and lymphoma (Phase I dose-escalation data expected H1 2026), and the recently dosed REC-3565, a MALT1 inhibitor for B-cell lymphomas. Integration has not been without friction — several legacy candidates were deprioritized post-merger — but the combined platform represents the closest thing to an end-to-end AI drug development operating system.

What to watch: Phase II data for REC-3964 and Phase I escalation data for REC-1245 will be the first clinical tests of the merged platform’s output.

3. Schrödinger

Headquarters: New York, NY | Founded: 1990 Platform: Physics-based molecular simulation + machine learning Most Advanced Asset: Zasocitinib (TAK-279) — Phase III (via Takeda partnership)

Schrödinger’s physics-based computational platform has produced what is currently the most clinically advanced AI-assisted drug in the world. Zasocitinib, a TYK2 inhibitor originally discovered by Nimbus Therapeutics using Schrödinger’s platform, is now in Phase III clinical trials through a partnership with Takeda for inflammatory and autoimmune diseases. Bristol Myers Squibb acquired the rights to the original Nimbus TYK2 program for $6 billion, validating the commercial potential of Schrödinger-enabled discovery.

Schrödinger’s approach differs from most AI drug discovery companies: rather than relying primarily on data-driven machine learning, their platform uses quantum mechanics-based molecular simulations to predict how drug molecules will behave. This physics-first approach has attracted collaborations with multiple large pharmaceutical companies.

What to watch: Phase III readouts for zasocitinib will be the highest-stakes clinical test of AI-enabled drug design to date.

4. Relay Therapeutics

Headquarters: Cambridge, MA | Founded: 2016 Platform: Dynamo (AI-driven protein motion analysis for precision medicine) Most Advanced Asset: Reltlimab combination studies, RLY-2608 (PI3Kα inhibitor, Phase I)

Relay Therapeutics uses AI to understand protein motion — how proteins change shape and move in real time — to design drugs that exploit those dynamics. Their lead program RLY-2608 is a mutant-selective PI3Kα inhibitor being developed for breast cancer and other solid tumors. The approach targets a specific mutation while sparing the normal protein, a level of precision that has historically been extremely difficult to achieve with traditional drug design.

What to watch: Clinical data expansion for RLY-2608 across additional tumor types.

5. AbCellera Biologics

Headquarters: Vancouver, Canada | Founded: 2012 Platform: AI-powered antibody discovery at massive scale Clinical Validation: Multiple partnered programs in the clinic

AbCellera’s platform combines microfluidics, high-throughput screening, and AI to discover antibodies from natural immune responses. Their technology was famously used to discover bamlanivimab, which received emergency use authorization for COVID-19 in just 90 days from sample to clinical use. The company has active partnerships with more than 170 pharmaceutical and biotech companies, with over 200 discovery programs initiated across therapeutic areas including oncology, immunology, neuroscience, and infectious disease. Multiple partnered programs have advanced into clinical development.

What to watch: Internal pipeline advancement and new partnership disclosures.

6. BenevolentAI

Headquarters: London, UK | Founded: 2013 Platform: Knowledge graph + machine learning for target discovery and drug repurposing Most Advanced Asset: BEN-8744 (Phase II for ulcerative colitis)

BenevolentAI’s platform maps complex biological relationships by mining scientific literature, patents, clinical data, and omics datasets to identify novel drug targets and repurposing opportunities. Their lead program BEN-8744, a PDE10 inhibitor for ulcerative colitis, was identified through the platform’s ability to discover non-obvious connections between targets and diseases. The company gained global attention when its AI platform identified baricitinib as a potential COVID-19 treatment, which was subsequently validated in clinical trials.

What to watch: Phase II data for BEN-8744 will be a key test of the knowledge graph approach to target discovery.

Tier 2: Platform Validated Through Major Pharma Partnerships

These companies have demonstrated platform credibility through significant pharmaceutical partnerships, with clinical candidates in development.

7. Chai Discovery

Headquarters: San Francisco, CA | Founded: 2024 Platform: AI-driven biologics design Key Partnerships: Eli Lilly (2026)

Chai Discovery may be the fastest-growing company in AI drug discovery history. Eighteen months after launch, the company closed a $130 million Series B at a $1.3 billion valuation. Their January 2026 partnership with Eli Lilly is structurally different from typical pharma-AI deals: Chai is deploying its technology across multiple biologics targets for Lilly while building an exclusive AI model trained on Lilly’s proprietary data. This represents a new partnership model — not a single-asset collaboration, but an infrastructure integration. Co-founder Jack Dent has stated publicly that he believes 2026 will be the year AI drug discovery moves from research breakthroughs to deployment.

What to watch: First disclosed clinical candidates from the Lilly partnership.

8. Isomorphic Labs

Headquarters: London, UK | Founded: 2021 Platform: AI-first drug design built on AlphaFold (Google DeepMind spinout) Key Partnerships: Eli Lilly, Novartis

Isomorphic Labs was born from the team behind AlphaFold, the AI system that solved protein structure prediction and won the 2024 Nobel Prize in Chemistry. Their partnerships with Eli Lilly and Novartis — collectively valued at up to $3 billion — deploy Isomorphic’s AI to design small molecule and biologic drug candidates across multiple therapeutic areas. While no Isomorphic-originated drugs have reached clinical trials yet, the depth of their pharma relationships and the pedigree of their underlying technology make them one of the most closely watched companies in the space.

What to watch: Transition from preclinical candidates to IND filings over the next 12 to 18 months.

9. Atomwise

Headquarters: San Francisco, CA | Founded: 2012 Platform: AtomNet (structure-based deep learning for small molecule discovery) Key Partnerships: 250+ research institutions, multiple pharma collaborations

Atomwise’s AtomNet platform uses convolutional neural networks to predict how small molecules bind to protein targets, screening billions of virtual compounds in the time traditional methods would screen thousands. The company has more than 250 active collaborations across oncology, neurology, and infectious disease. Multiple Atomwise-discovered compounds are in preclinical and clinical development through partner programs.

What to watch: Advancement of partnered programs into later-stage clinical trials.

10. Noetik

Headquarters: San Francisco, CA | Founded: 2023 Platform: AI for cancer clinical outcome prediction Key Partnerships: GSK (2026)

Noetik’s early 2026 partnership with GSK focuses on using AI to predict clinical outcomes in oncology — a different application than most AI drug discovery companies, which focus on molecule design. By predicting which patients are most likely to respond to specific therapies, Noetik’s platform could reshape clinical trial design and patient selection in cancer. The GSK deal signals major pharma confidence in AI for clinical decision-making, not just molecular design.

What to watch: Integration into GSK’s oncology clinical trial design process and first trial results using Noetik-guided patient selection.

11. Boltz

Headquarters: (Stealth/early-stage) | Founded: Recent Platform: AI for small molecule drug discovery Key Partnerships: Pfizer (2026)

Boltz rounds out the January 2026 trifecta of major pharma-AI platform deals. Their partnership with Pfizer deploys Boltz’s technology across small molecule drug discovery programs. Details remain limited, but Pfizer’s willingness to sign a platform deal — rather than a single-asset collaboration — signals confidence in the underlying technology.

What to watch: Disclosed program details and preclinical candidate nominations from the Pfizer partnership.

12. Iambic Therapeutics

Headquarters: San Diego, CA | Founded: 2021 Platform: Physics-based AI + high-throughput experimentation Key Partnerships: Takeda, Eli Lilly

Iambic’s platform combines physics-based AI algorithms with a high-throughput experimental lab that converts new molecular designs into biological data on a weekly cycle. Their collaboration with Takeda targets some of the most challenging molecular design problems in drug discovery. Iambic emphasizes putting AI tools in the hands of experienced medicinal chemists, rather than replacing them — a philosophy that has attracted multiple pharmaceutical partnerships.

What to watch: Clinical candidate nominations from the Takeda and Lilly partnerships.

13. XtalPi

Headquarters: Shenzhen, China / Cambridge, MA | Founded: 2014 Platform: ID4 (quantum mechanics + AI for molecular property prediction) Key Partnerships: Pfizer, multiple pharmaceutical companies

XtalPi’s Intelligent Digital Drug Discovery and Development platform uses quantum chemistry and AI to predict molecular properties like solubility, stability, and crystal forms with high precision. Crystal form prediction is a critical and often underappreciated step in drug development — the wrong crystal form can render a promising molecule unmanufacturable. XtalPi’s ability to predict this computationally has attracted pharmaceutical partners worldwide.

What to watch: Expansion of pharmaceutical partnerships and clinical advancement of XtalPi-enabled compounds.

Tier 3: Advancing Platform Companies With Strong Pipeline Momentum

These companies have demonstrated significant platform capabilities and are rapidly advancing toward or entering clinical development.

14. Nimbus Therapeutics

Headquarters: Cambridge, MA | Founded: 2009 Platform: Structure-based computational drug design Key Validation: $6 billion acquisition of TYK2 program by Bristol Myers Squibb

Nimbus is the original proof that computationally designed drugs can command blockbuster valuations. Their TYK2 inhibitor program, designed using Schrödinger’s computational platform, was acquired by Bristol Myers Squibb for up to $6 billion. Nimbus continues to advance a pipeline of computationally designed programs across metabolic and immunological diseases.

What to watch: Next-generation pipeline programs and potential further asset deals.

15. Generate Biomedicines

Headquarters: Somerville, MA | Founded: 2018 Platform: Generative AI for protein therapeutics design Key Partnerships: Novartis ($1 billion+ deal), Amgen

Generate Biomedicines uses generative AI to design novel protein therapeutics from scratch — not just optimizing existing proteins, but creating entirely new ones. Their partnership with Novartis, valued at over $1 billion, targets the design of protein-based therapies across multiple disease areas. A separate collaboration with Amgen extends the platform to additional therapeutic modalities.

What to watch: First clinical candidates from the Novartis partnership.

16. Evotec

Headquarters: Hamburg, Germany | Founded: 1993 Platform: Integrated AI + wet lab drug discovery platform Key Partnerships: Bayer, Bristol Myers Squibb, multiple pharma companies

Evotec operates one of the largest integrated drug discovery and development platforms in the world, combining AI-driven computational tools with massive wet lab capabilities. Their partnerships with Bayer, Bristol Myers Squibb, and others span target discovery through clinical development. Evotec’s scale — thousands of scientists across multiple sites — gives it an advantage in generating the high-quality proprietary data that AI models need to perform well.

What to watch: Advancement of AI-enabled pipeline programs toward clinical trials.

17. Absci Corporation

Headquarters: Vancouver, WA | Founded: 2011 Platform: Generative AI for de novo antibody design Key Capability: “Zero-shot” antibody generation without extensive training data

Absci’s Integrated Drug Creation platform uses generative AI to design novel therapeutic antibodies from scratch, including their landmark “zero-shot” approach that can create functional antibodies without requiring extensive existing data on the target. This capability addresses one of the fundamental limitations in AI drug design — the dependence on large historical datasets.

What to watch: Clinical advancement of internally discovered antibody candidates and new partnership announcements.

18. Tempus AI

Headquarters: Chicago, IL | Founded: 2015 Platform: Multimodal clinical and molecular data platform Key Capability: AI-powered precision medicine across 350+ petabytes of clinical data

Tempus operates at the intersection of AI and clinical data. With over 350 petabytes of de-identified clinical and molecular data, Tempus powers precision medicine by helping match patients to optimal therapies and clinical trials. While not a traditional drug discovery company, Tempus’s data platform is increasingly being used by pharmaceutical companies to identify novel targets, design clinical trials, and stratify patient populations — making it critical infrastructure for AI-enabled drug development.

What to watch: Expansion of pharmaceutical R&D partnerships and potential internal pipeline development.

19. Owkin

Headquarters: Paris, France / New York, NY | Founded: 2016 Platform: Privacy-preserving federated AI for drug development Key Partnerships: Sanofi ($180 million+ deal), multiple academic medical centers

Owkin’s federated learning approach allows AI models to train on data from multiple hospitals without that data ever leaving the institution — solving one of the biggest barriers in healthcare AI: data access and privacy. Their partnership with Sanofi spans oncology, immunology, and inflammation, using Owkin’s technology to identify biomarkers and patient populations. Owkin’s approach to the data access problem could prove to be as important as any molecule design platform.

What to watch: Biomarker discoveries and clinical trial applications emerging from the Sanofi partnership.

20. Valo Health

Headquarters: Boston, MA | Founded: 2019 Platform: Opal (computational platform integrating human-centric data with AI) Most Advanced Asset: VHD-3482 (oncology, entering clinical development)

Valo Health’s platform integrates patient-level health data with computational chemistry and machine learning to discover and develop drugs that are optimized for both biological activity and clinical success. Their approach emphasizes using real-world human data early in the discovery process to increase the probability of clinical translation — addressing the persistent problem that drugs which work in preclinical models often fail in humans.

What to watch: Clinical entry and first data from VHD-3482.

21. Verge Genomics

Headquarters: South San Francisco, CA | Founded: 2015 Platform: Human-centric AI for neuroscience drug discovery Most Advanced Asset: VRG50635 (ALS, Phase I completed)

Verge Genomics focuses exclusively on neuroscience — one of the most difficult therapeutic areas for drug development, with failure rates above 99% in Alzheimer’s disease. Their platform uses AI to analyze human brain tissue data rather than animal models, identifying drug targets directly from human disease biology. VRG50635, their lead program for ALS, has completed Phase I dosing. The neuroscience focus is deliberate: if AI can improve success rates in the hardest therapeutic area, the platform’s value is proven beyond doubt.

What to watch: Phase I data from VRG50635 and next-stage clinical development decisions.

22. Healx

Headquarters: Cambridge, UK | Founded: 2014 Platform: AI for rare disease drug repurposing Key Capability: Identification of existing drugs that can treat rare diseases

Healx uses AI to find new therapeutic uses for existing approved drugs — specifically targeting rare diseases, where the economics of traditional drug development rarely work. Their platform has identified multiple repurposing candidates, several of which are in preclinical or clinical-stage development. The rare disease focus creates a unique position: shorter development timelines (since the drug’s safety profile is already established), orphan drug incentives, and a deeply underserved patient population.

What to watch: Clinical advancement of repurposed drug candidates in rare disease indications.

23. BigHat Biosciences

Headquarters: San Mateo, CA | Founded: 2019 Platform: AI + high-throughput experimentation for antibody engineering Key Partnerships: Merck, multiple undisclosed pharma partners

BigHat combines machine learning with a high-throughput wet lab to engineer antibodies with optimized properties — affinity, stability, manufacturability, and developability. Their partnership with Merck and multiple undisclosed collaborations validate the platform’s utility for large pharmaceutical companies looking to optimize their biologic pipelines.

What to watch: Advancement of Merck-partnered and internal programs toward clinical development.

24. Exscientia (Legacy Programs)

Headquarters: Oxford, UK | Founded: 2012 Platform: AI-driven precision molecular design (now integrated into Recursion) Key Legacy: First AI-designed drug to enter clinical trials (2020)

While Exscientia is now part of Recursion (covered in Tier 1), its legacy deserves separate recognition. Exscientia was the first company to put an AI-designed drug into clinical trials, in 2020 — a landmark moment for the field. Their precision chemistry platform, partnerships with Sanofi, Bristol Myers Squibb, and Merck KGaA, and multiple clinical-stage programs established the template that other AI drug discovery companies have followed. The Exscientia approach of automating the design-make-test-analyze cycle is now core to the merged Recursion operating system.

25. PeptiDream

Headquarters: Kawasaki, Japan | Founded: 2006 Platform: Peptide Discovery Platform System (PDPS) + computational design Key Capability: Discovery of macrocyclic peptide therapeutics using AI and computational modeling

PeptiDream’s platform uses advanced computational modeling and proprietary screening technology to discover peptide-based drug candidates. Macrocyclic peptides occupy a unique chemical space between small molecules and large biologics, enabling them to hit targets that neither traditional modality can effectively reach. PeptiDream has active partnerships with more than 20 pharmaceutical companies and an extensive royalty-bearing portfolio.

What to watch: Advancement of partnered pipeline programs and expansion of computational capabilities.

The Landscape in Numbers

A few data points that put this list in context:

The number of AI-originated drugs entering clinical development has grown exponentially, from 3 in 2016 to 67 in 2023 to more than 200 as of early 2026. Phase I success rates for AI-discovered compounds run between 80% and 90%, compared to approximately 40% to 65% for traditionally discovered drugs. AI is compressing the discovery-to-clinic timeline from a historical average of 4 to 6 years to as little as 18 months in some cases, and is estimated to reduce preclinical costs by 30% to 70%.

The January 2026 platform deals between Eli Lilly and Chai Discovery, GSK and Noetik, and Pfizer and Boltz collectively signal that major pharmaceutical companies now view AI not as an experiment but as core R&D infrastructure. Eighty-one percent of pharmaceutical companies report deploying AI in some capacity.

But the critical statistic remains unresolved: no AI-designed drug has yet completed a Phase III trial and received regulatory approval. That milestone — expected by many analysts to occur in 2026 or 2027 — will be the definitive proof point for the entire field.

What Separates the Real From the Hype

Not every company calling itself an “AI drug discovery company” belongs on this list. Here are the signals that distinguish credible platforms from noise:

Clinical assets in development. The ultimate proof of an AI platform is a molecule in human trials. Companies that have been operating for 5+ years without advancing a clinical candidate deserve scrutiny.

Pharmaceutical partnerships with disclosed economics. When Eli Lilly pays for an exclusive AI model or Sanofi commits $180 million to a collaboration, those are credibility signals that no amount of conference presentations can replicate.

Proprietary data generation. The companies that generate their own experimental data — through automated labs, clinical data platforms, or federated learning networks — have a structural advantage over those relying on public datasets.

Therapeutic area focus vs. platform breadth. Both models can work, but the most credible companies can articulate clearly where their AI adds the most value and where it doesn’t.

Honest communication about limitations. AI cannot solve fundamental biological complexity. Companies that acknowledge this and position AI as a tool that augments human scientists — rather than a replacement — tend to produce better clinical outcomes.

Who’s Missing — and Why

This list deliberately excludes several categories of companies. Large pharmaceutical companies with internal AI capabilities (Pfizer, Roche, AstraZeneca) are using AI extensively but are not “AI drug discovery companies” in the same sense. Pure-play computational chemistry companies that predate the current AI wave (many academic spinouts) were excluded unless their platforms have been substantially enhanced with modern machine learning. And early-stage companies without disclosed clinical programs or major partnerships, regardless of the quality of their science, did not meet the clinical evidence filter.

We recognize this creates blind spots. Breakthrough platforms may be operating in stealth. Academic spinouts may be generating extraordinary science that hasn’t yet translated into industry partnerships. We welcome nominations for the next quarterly update.

Is your company advancing AI-designed therapeutics into clinical development? Contact our editorial team for consideration in the next update.

What Happens Next

The next 12 months will be the most consequential in the history of AI drug discovery. Phase III readouts for zasocitinib, continued data from rentosertib, and multiple Phase II results from across this list will collectively answer the question that has defined the field for a decade: does AI actually make better drugs, or does it just make them faster?

If the data is positive, the companies on this list will become some of the most valuable platforms in biotechnology. If the data disappoints, the industry faces a fundamental recalibration.

Either way, the answer starts here.

Subscribe to BioMed Nexus for daily coverage of AI drug discovery, clinical readouts, and the deals reshaping pharmaceutical R&D.

Last updated: March 2026. Next update: June 2026.