AI Drug Discovery: what are the top startups now?

Which AI drug discovery startups are getting the biggest fresh checks right now?

Isomorphic Labs is clearly ahead on raw capital, while Earendil Labs, Chai Discovery, Profluent, Iambic, Medra, Chemify, and Genesis Therapeutics form the fresher emerging pack. The difference is important. Isomorphic’s round is in a different league, but the smaller rounds tell us where the market is warming up next.

The biggest signal, by far, is Isomorphic Labs. In May 2026, the company announced a $2.1 billion Series B led by Thrive Capital, with Alphabet, GV, MGX, Temasek, CapitalG, and the UK Sovereign AI Fund involved. In AI drug discovery, that is not normal biotech financing. It is closer to infrastructure-company money. Investors are basically saying: this may become a core AI layer for drug design, not just another biotech pipeline.

That puts Isomorphic above everyone else on fresh capital. To compare: Chai Discovery raised $130 million Series B in December 2025 at a $1.3 billion valuation, which is already huge for a young AI biology company. But Isomorphic raised more than 16 times that amount in one round. So Chai is extremely hot, but Isomorphic is operating at a different funding scale.

The interesting part is what comes after Isomorphic. Earendil Labs announced $787 million in financing in March 2026 for AI-driven biologics discovery and development. That is much bigger than Chai, Profluent, Iambic, Medra, or Chemify on headline size. But there is a catch: Earendil’s public proof is still thinner. We see the money and the biologics ambition, but we do not yet see enough public pipeline or partner evidence to say it is ahead of Chai in execution.

Chai Discovery is more balanced. Its round was smaller than Earendil’s, but it came with a clear model story and then, one month later, a Lilly collaboration for biologics discovery. That combination makes Chai easier to rank highly today. Earendil has the bigger financing headline. Chai has the cleaner public validation stack.

Profluent raised $106 million in November 2025, bringing total funding to about $150 million, for AI-designed proteins, genome editors, antibodies, and enzymes. Compared with Chai, Profluent feels less like a pharma-discovery partner today and more like a programmable-biology moonshot. That can be more technically exciting, but it is also harder to evaluate from outside.

Iambic raised more than $100 million in November 2025. It sits in a more practical lane than Profluent: AI-discovered therapeutics moving toward or into human trials. If we rank by “could this become a clinical biotech?”, Iambic looks stronger. If we rank by “could this change how biology is programmed?”, Profluent looks more ambitious.

Medra and Chemify are smaller but more interesting than their round sizes suggest. Medra raised $52 million Series A in December 2025 for “Physical AI Scientists,” meaning AI systems connected to real lab execution. Chemify raised more than $50 million Series B in October 2025 to scale automated digital chemistry. These are not the biggest checks, but they point to a very important shift: investors are now backing the lab loop, not only the model.

Finally, Genesis Therapeutics belongs in this capital bucket because of its May 2026 Incyte expansion. It was not a normal venture round, but Incyte committed $120 million in total upfront consideration, including $80 million cash and $40 million equity, plus recurring research funding. In some ways, that is a stronger signal than venture money because it comes from a pharma partner using the platform.

If you want more recent data on this point, please see our latest AI in drug discovery market report.

Which AI drug discovery startups are pharma companies actually paying?

Insilico Medicine and Genesis Therapeutics stand out most on recent paid pharma validation, with Chai Discovery, Formation Bio, Owkin, insitro, Isomorphic Labs, and Medra also showing strong partner signals. This is one of the best filters because pharma partnerships are harder to fake than broad claims about “AI-powered discovery.”

The strongest recent deal momentum belongs to Insilico Medicine. In March 2026, Lilly expanded its partnership with Insilico in a deal with $115 million upfront and up to $2.75 billion in milestones. Then in June 2026, SK Biopharmaceuticals signed a neuroimmune collaboration with Insilico worth up to $2.5 billion. The SK deal had only $18 million upfront and near-term milestones, so we should not treat the $2.5 billion as guaranteed money. But even with that caveat, two large pharma deals in the same year make Insilico one of the most commercially validated companies in the category right now.

Genesis Therapeutics is the other standout because the Incyte expansion is very tangible. In May 2026, Incyte agreed to $120 million in total upfront consideration, split between $80 million cash and $40 million equity, plus recurring research funding. That is not just a nice logo on a partnership slide. It shows a pharma partner was willing to expand the collaboration and commit serious near-term capital. Compared with Insilico, Genesis has less public clinical breadth, but its recent partner signal is cleaner than many larger headline milestone deals.

Chai Discovery is less proven economically because Lilly did not disclose the financial terms of its January 2026 collaboration. But the timing is powerful. Chai raised at a $1.3 billion valuation in December 2025, then announced Lilly access to its AI platform and custom models for biologics discovery in January 2026. If we only rank paid terms, Insilico and Genesis are ahead. If we rank “fresh pharma interest in a new AI biology model company,” Chai is one of the top names.

Formation Bio is important because it is playing a different game. It is not only trying to discover drugs faster but to run development differently. Sanofi licensed Formation’s phase 3-stage oral JAK/SYK inhibitor gusacitinib in a deal worth up to roughly €545 million, or about $626 million to $632 million depending on exchange rate and reporting. Formation also worked with Sanofi and OpenAI on Muse, an AI tool for clinical-trial recruitment. That makes Formation one of the few companies where the AI story goes beyond molecule design.

Owkin has one of the freshest pharma-workflow signals. In June 2026, Sanofi announced a multi-year collaboration with Owkin to co-develop biopharma AI agents, including a five-year license for Owkin’s K Pro “AI Scientist.” This is not the same kind of proof as a drug candidate entering the clinic. But it tells us pharma is now willing to bring AI agents inside R&D workflows, not just use AI as an external discovery service.

insitro is quieter than the newest model companies, but still relevant. In October 2025, Bristol Myers Squibb extended its ALS collaboration with insitro around ChemML-driven molecule discovery. In March 2026, insitro said the BMS collaboration had expanded with new target nominations. It also announced a Lilly collaboration in September 2025 for machine-learning models in small-molecule discovery. That is steady partner validation, even if it does not create the same buzz as Isomorphic or Chai.

Isomorphic Labs should be treated separately. It has major pharma relationships, but its freshest 2026 signal is the massive Series B. So in this specific category, we would put Insilico and Genesis above Isomorphic. Isomorphic may be the bigger company story, but Insilico and Genesis have more recent concrete pharma-payment evidence.

Medra is still early, but the Genentech partnership attached to its December 2025 Series A matters. Autonomous lab companies need real pharma workflows to prove they are not just robotics demos. Medra has not yet proven scale, but it is working on the right problem with the right type of partner.

Which AI drug discovery startups are closest to proving AI can create real drugs?

Recursion and Insilico Medicine are ahead on clinical proof, while Iambic and CHARM Therapeutics are the more interesting private challengers. Isomorphic and Chai may be hotter on model ambition, but they are not yet as easy to judge on public clinical evidence.

This is where Recursion still matters a lot. The company is public, so we get more visibility than with most private AI drug discovery startups. In May 2026, Recursion reported early clinical data for REC-1245, Phase 2 efficacy signals for REC-4881, and first patient dosing for REC-4539. That is the type of evidence we should care about: programs, patients, readouts, and FDA-facing development work. Recursion is no longer just asking people to believe in scaled biology. It is being judged by the same brutal standard as every biotech: do the drugs work?

Insilico Medicine is the other clinical-proof leader. Its lead IPF candidate, ISM001-055, is one of the best-known examples of an AI-designed drug reaching clinical trials. More importantly, Insilico now has breadth. The company has said it has 28 AI-developed drugs, with about half in clinical trials, and recent reporting said it had 13 drug candidates cleared or approved in China, including 10 in clinical trials. Even if we treat company-reported pipeline numbers carefully, Insilico is clearly beyond the “cool platform demo” stage.

Compared with Recursion, Insilico looks stronger on the “AI-designed asset” narrative. Compared with Insilico, Recursion looks stronger on public-market transparency and scaled experimental biology. That distinction matters. Recursion is easier to inspect. Insilico is easier to point to as a pure AI-generated drug story.

Iambic is one of the private companies worth watching here. In November 2025, it said its platform had delivered new drug candidates into human clinical trials across multiple target classes and mechanisms, and it raised more than $100 million to keep advancing AI-discovered therapeutics. We do not yet have the same public clinical depth as Recursion, but among private startups, Iambic looks meaningfully closer to clinical proof than most model-first companies.

CHARM Therapeutics is narrower but concrete. In September 2025, it raised $80 million Series B to move a next-generation menin inhibitor into clinical development. That is not as broad as Recursion or Insilico, but it is easier to understand than many AI drug discovery claims: one serious oncology direction, a defined asset strategy, and a clear push into the clinic.

Isomorphic Labs is tricky in this section. It has the strongest capital signal in the entire market, and its DeepMind roots make it impossible to ignore. But if the question is “who has already shown the most public clinical proof?”, Isomorphic does not beat Recursion or Insilico today. It may become the most important company in the market. It just has less visible clinical evidence right now.

Same for Chai Discovery. Chai is extremely hot as a model and biologics platform, but we cannot yet rank it above Recursion, Insilico, or Iambic for clinical proof. A Lilly collaboration is a strong validation signal. It is not the same as human efficacy data.

Recursion and Insilico are the clinical-proof leaders, Iambic is the private challenger to watch, CHARM is a focused oncology entrant, and Isomorphic/Chai still need more public clinical evidence before they can win this specific category.

If you want more recent data on this point, please see our latest AI in drug discovery market report.

Which startups are leading the foundation-model shift in biology?

Isomorphic Labs and Chai Discovery are the clearest leaders now, while EvolutionaryScale, Profluent, Xaira, and Owkin each matter for a different reason. This is where the market has changed the most. A few years ago, many AI drug discovery companies were basically saying: “we can screen molecules faster.” Today, the more exciting companies are saying: “we can build reusable models of biology.”

Isomorphic Labs is the flagship. It has the capital, the DeepMind connection, and the ambition to turn AI into a general drug-design engine. Its May 2026 Series B makes more sense if we view Isomorphic as a biology-model infrastructure company rather than a normal biotech. A classic biotech does not usually raise $2.1 billion before becoming a mature commercial drug company. A potential AI platform layer can.

Chai Discovery is the strongest emerging challenger because it has a cleaner “new foundation model company” profile than most. The company is focused on molecular structure prediction, molecular interaction modeling, and computer-aided design. Its December 2025 valuation and January 2026 Lilly collaboration make the story feel current, not theoretical. Compared with Isomorphic, Chai is smaller and earlier. Compared with most new AI biology startups, it has better public validation.

EvolutionaryScale is important, but in a different lane. It is closer to the model layer than to a classic drug company. Its ESM models are influential for protein representation and protein generation. The reason we do not put it above Chai is commercial proof. EvolutionaryScale may be technically important, but Chai has the fresher pharma-discovery validation.

Profluent is probably the most interesting moonshot here. It is applying AI to programmable biology: proteins, genome editors, antibodies, and enzymes. If Chai feels like a drug-discovery model company, Profluent feels more like a company trying to make biology writable. That could be huge, but it is also harder to benchmark from outside because the commercial path is less direct.

Xaira is the one we have to handle carefully. It launched with more than $1 billion in committed capital and an elite scientific team. That alone makes it one of the most important AI biology companies. But compared with Isomorphic or Chai, Xaira has less recent public proof tied to specific pharma partnerships, products, or clinical progress. So it belongs in the top-watch list, but with a proof discount.

Owkin is not a foundation-model biology company in the same way as Isomorphic, Chai, or EvolutionaryScale. But its June 2026 Sanofi collaboration around K Pro and AI agents shows another version of the same trend: pharma wants AI systems that can reason across complex biological and clinical data, not just run a narrow prediction task.

Which startups are strongest in AI-designed biologics and protein engineering?

Chai Discovery currently looks like the cleanest leader in AI biologics, while Earendil Labs has the biggest new-money signal and Profluent has the most ambitious programmable-biology angle. EvolutionaryScale, Cradle, and Xaira are also relevant, but they do not all lead on the same metric.

Chai Discovery is first because its evidence is balanced. It has a fresh valuation, a clear biologics-model thesis, and a named pharma collaboration with Lilly. That combination beats companies that only have capital or only have technical claims. Chai may not have clinical proof yet, but for AI-designed biologics today, it has one of the strongest public validation stacks.

Earendil Labs is the wild card. Its $787 million March 2026 financing is massive for an AI biologics company. If we ranked only by fresh capital in biologics, Earendil would be first. But if we rank by public evidence we can compare, Chai is easier to defend. Earendil may be building something very serious, but the outside world still has fewer concrete data points on pipeline quality, pharma usage, or technical benchmarks.

Profluent is different again. They are trying to design proteins and biological tools, including genome editors and enzymes. That makes Profluent more technically radical than Chai. But Chai looks more directly connected to pharma drug-discovery demand today. So the ranking depends on the question: for near-term AI biologics discovery, Chai leads; for broader programmable biology, Profluent may be the more exciting bet.

EvolutionaryScale matters because the ESM model family is one of the major technical references in protein modeling. But we should separate model influence from company traction. EvolutionaryScale is important to the field. Chai currently has stronger evidence as a startup that pharma is using for drug discovery.

Cradle is a more productized protein-engineering company. It gives scientists tools to design and improve proteins with AI and wet-lab feedback. That makes it practical and potentially useful, but it does not have the same blockbuster signal as Chai’s Lilly collaboration, Earendil’s financing, or Profluent’s programmable-biology ambition.

Xaira belongs because of its scale, capital, and scientific team. But again, compared with Chai, it has less public evidence we can point to right now. Xaira may become one of the winners. Today, it is still more of a very well-funded promise than a fully benchmarkable leader.

If you want more recent data on this point, please see our latest AI in drug discovery market report.

Which startups are strongest in AI small-molecule discovery?

Insilico Medicine, Recursion, Genesis Therapeutics, and Iambic are the strongest small-molecule AI names right now, while CHARM Therapeutics, Terray Therapeutics, and Chemify fill more specific roles. This is a more mature category than AI biologics, so the bar is higher. A company should not get credit just because it says it uses AI to find molecules.

Insilico Medicine is the clearest AI small-molecule leader if we care about the full chain from platform to clinical assets to pharma deals. It has AI-developed clinical candidates, a broad pipeline, and fresh 2026 pharma deal activity. Compared with newer model companies, Insilico looks less fashionable. Compared with most of them, it has much more proof that its platform can generate drug candidates that enter real development.

Recursion is the other major leader, but with a different flavor. Recursion is not only about designing one molecule at a time. It is about scaled biology, automated experiments, image-based phenomics, machine learning, and chemistry after the Exscientia combination. Compared with Insilico, Recursion has more visible public-company reporting and infrastructure. It is less cleanly associated with one famous AI-designed drug story.

Genesis Therapeutics is the most interesting private small-molecule name in the current evidence set. The Incyte expansion suggests repeat use of the platform across multiple targets, not just a one-time pilot. That is why Genesis deserves to be ranked above many AI discovery startups with louder branding but weaker recent partner economics.

Iambic is the private clinical challenger. It has raised serious money and says it has delivered candidates into human trials across multiple target classes. We would not put it above Insilico or Recursion yet, because those companies have more visible breadth and public evidence. But Iambic is one of the few private names that feels meaningfully close to the clinical frontier.

CHARM Therapeutics is focused and therefore easier to judge. Its menin-inhibitor direction gives it a concrete oncology story. It does not look like a broad AI platform leader yet, but it is more real than companies that only talk about discovery speed without naming assets.

Terray Therapeutics remains relevant, although its biggest financing signal is older than the freshest 2025 and 2026 rounds. Terray’s strength is the integration of computation with novel data at scale for small molecules. The reason it is not in the top four here is recency. The thesis is strong, but the latest public heat is stronger around Insilico, Genesis, and Iambic.

Chemify is the infrastructure name in this small-molecule group. It is less about choosing the best molecule on a screen and more about designing and making molecules through automated chemistry. That matters because synthesis is a real bottleneck. A generated molecule is not very useful if nobody can make it quickly and reliably.

In small molecules, Insilico and Recursion remain the serious leaders, Genesis is the freshest partner-validated private name, Iambic is the clinical-stage private challenger, CHARM is a focused oncology entrant, Terray is still credible but less fresh, and Chemify is the picks-and-shovels chemistry bet.

Which startups are building the self-driving lab for drug discovery?

Medra and Chemify are the freshest pure plays, while Recursion, Terray Therapeutics, and Cradle show different versions of the same closed-loop idea. This category may be less flashy than foundation models, but it could become more important over time. The reason is simple: biology needs experiments. A model that cannot learn from real lab feedback will eventually hit a wall.

Medra is the freshest name here. In December 2025, it raised $52 million Series A to build “Physical AI Scientists” and announced a Genentech partnership. The phrase sounds futuristic, but the concept is very concrete: connect AI planning to real robotic lab execution. Compared with Recursion, Medra is much earlier. Compared with most new AI biology companies, it is attacking a more practical bottleneck.

Chemify is the strongest chemistry-automation name. Its October 2025 Series B was about scaling digital chemistry and automated Chemifarm facilities. If Medra is trying to automate the scientist, Chemify is trying to digitize and automate the chemistry process. That makes Chemify especially relevant for small-molecule discovery.

Recursion has the most mature scaled infrastructure in this group. It has spent years building high-throughput biological data generation and computational systems. The reason we do not call it a pure self-driving-lab startup is that it is broader than that. But if the question is “who already has a large experimental loop?”, Recursion is ahead of Medra and Chemify.

Terray Therapeutics also belongs because its platform combines computation with novel data generation for small molecules. It is less fresh in the headlines than Medra and Chemify, but the underlying idea is close: build a data engine that improves discovery through repeated experiment and model feedback.

Cradle is a narrower protein-engineering version of the loop. Scientists use the software, experiments generate feedback, and the model improves protein design. It does not look like a full autonomous lab company today, but it fits the broader direction.

Which startups are using AI to improve clinical development, not just discovery?

Formation Bio leads this category, with Owkin, Recursion, Insilico Medicine, and insitro also standing out. This angle matters because discovering a molecule is only one part of the problem. Many drugs fail because of trial design, patient selection, recruitment delays, biology mismatch, toxicity, or weak development execution.

Formation Bio is the cleanest leader here because its whole positioning is AI-native drug development. The Sanofi gusacitinib deal gives it a serious late-stage asset signal, and the Muse work with Sanofi and OpenAI shows a practical use case: using AI for clinical-trial recruitment. Compared with Isomorphic or Chai, Formation is less glamorous. Compared with them, it is closer to the messy development bottlenecks where pharma spends real money today.

Owkin is also strong because it has long focused on patient data, biology, and pharma workflows. Its June 2026 Sanofi collaboration around K Pro and AI agents makes it one of the better current examples of AI moving inside R&D decision-making. Owkin is not winning because it designed the most famous molecule but because pharma needs better ways to reason across clinical and biological data.

Recursion belongs here because it now has to execute clinical development, not only discovery. Once a company has multiple programs in the clinic, AI claims become less abstract. The question becomes: can the platform pick better programs, design better trials, and kill weak ones faster?

Insilico Medicine also belongs because it has a broad AI-developed pipeline and recent pharma demand. But compared with Formation, Insilico is still more strongly associated with discovery and asset generation. Formation is more directly positioned around development execution.

insitro is relevant because of its disease-biology and target-selection work, especially in genetically informed disease areas. The BMS and Lilly collaborations suggest pharma values its models for choosing and understanding targets, not just generating chemistry.

The ranking here is different from the overall AI drug discovery ranking. Formation Bio is first for clinical-development AI, Owkin is the strongest AI-agent/workflow name, Recursion and Insilico bring the clinical-pipeline angle, and insitro remains strong in target biology. This is probably where near-term ROI may show up before the first fully AI-designed blockbuster exists.

If you want more recent data on this point, please see our latest AI in drug discovery market report.

Which AI drug discovery startups look hottest but still need a proof discount?

Xaira, Earendil Labs, EvolutionaryScale, and Chai Discovery all deserve attention, but not all the same level of confidence. This does not mean they are weak companies. It means their public proof is not yet as strong as their ambition, funding, or technical narrative.

Xaira is the clearest example. The company launched with more than $1 billion in committed capital and an unusually strong team. That alone makes it impossible to ignore. But compared with Recursion or Insilico, we have much less public clinical proof. Compared with Chai, we have less fresh named pharma-discovery validation. So Xaira is a top-tier watchlist company, but not yet a top-tier evidence company.

Earendil Labs also needs a proof discount. Its $787 million financing is one of the biggest recent signals in AI biologics. But right now, the public information is still not rich enough to compare it properly with Chai on platform performance, pharma usage, or pipeline progress. Earendil may become a major winner. Today, the money is clearer than the proof.

EvolutionaryScale is technically important, especially because of the ESM model family. But being important to the protein-modeling ecosystem is not the same as proving leadership in drug discovery. If we rank technical influence, EvolutionaryScale is high. If we rank pharma traction or clinical progress, it is less proven.

Chai Discovery is the strongest of this “proof discount” group because it already has a Lilly collaboration and a fresh valuation. Still, it should not be treated like Recursion or Insilico yet. Chai has model and partner validation. It does not yet have public clinical proof that its designed molecules become successful medicines.

Which older AI drug discovery leaders are still genuinely relevant?

Recursion, Insilico Medicine, insitro, Owkin, and Genesis Therapeutics are still relevant, while some older virtual-screening names feel less central to the current heat. This question matters because AI drug discovery has been hyped for years. Some famous names still deserve attention. Others are mainly surviving on old reputation.

Recursion is still one of the category-defining companies because it has public clinical updates, the Exscientia combination, and a scaled biology platform. It is no longer the newest story, but it remains one of the few AI drug discovery companies we can judge with real public data.

Insilico Medicine is still very relevant because it keeps producing fresh signals. As seen above, its 2026 Lilly and SK Biopharmaceuticals deals put it near the top on pharma validation. Its clinical pipeline also makes it more concrete than many newer model-first companies.

insitro is less loud, but still serious. Its BMS and Lilly collaborations show continued pharma interest. The company’s disease-biology angle may not create the same buzz as generative protein design, but it remains valuable if it helps pharma pick better targets.

Owkin has stayed relevant by moving with the market. It started with patient-data and federated-learning credibility, and now it is pushing AI agents for biopharma R&D through the Sanofi collaboration. That is a good example of an older AI biotech adapting to the current agentic-AI wave.

Genesis Therapeutics is the older name that looks newly hot. The May 2026 Incyte expansion gives it a very fresh reason to be back in the conversation. Without that signal, it might have looked like one more AI small-molecule company. With it, Genesis becomes one of the most interesting private names in the market.

The older names that look less central are the ones still mostly associated with generic AI screening or older discovery workflows. That does not mean they are dead. It just means the market has moved. Today, the strongest companies either have clinical assets, repeated pharma use, major model-layer ambition, or proprietary experimental loops.

Which startups would we actually put in the top tier now?

The top tier today is Isomorphic Labs, Insilico Medicine, Recursion, Chai Discovery, and Genesis Therapeutics. They do not all win the same way, but they appear most often across the strongest evidence categories.

The next group is strong but more specialized. Formation Bio leads AI-native development. Owkin leads pharma AI-agent workflows. Iambic is one of the best private clinical-stage challengers. Profluent is the bold programmable-biology bet. Earendil Labs is the giant new biologics-financing signal. Medra and Chemify are the most interesting lab-automation and chemistry-automation names. Xaira has huge potential but still needs more public proof.

Bring everything together, and AI drug discovery leadership now depends on the loop, not just the model. The best companies either have pharma money, clinical assets, proprietary data generation, biological foundation models, or a way to connect AI predictions to real experiments. The startups that combine several of those signals are the ones that look like the real leaders today.

If you want more recent data on this point, please see our latest AI in drug discovery market report.

Category	Startups selected and why
Fresh capital leaders	Isomorphic Labs leads by far with its $2.1B Series B. Earendil Labs has the biggest new biologics financing. Chai Discovery is the cleanest emerging model-company signal. Profluent, Iambic, Medra, Chemify, and Genesis show where capital is moving next.
Pharma-paid validation	Insilico Medicine leads on recent pharma deal volume. Genesis Therapeutics has the cleanest recent private-company upfront signal. Formation Bio stands out in development-stage pharma work. Chai, Owkin, insitro, Isomorphic, and Medra add partner validation in more specific lanes.
Closest to clinical proof	Recursion and Insilico Medicine are ahead because they have visible clinical programs. Iambic is the private challenger. CHARM Therapeutics is focused but concrete. Isomorphic and Chai are hot, but less proven clinically today.
Foundation-model biology	Isomorphic Labs leads on scale. Chai Discovery leads among fresh emerging model companies. EvolutionaryScale matters technically. Profluent is the programmable-biology moonshot. Xaira has huge ambition but needs more public proof.
AI biologics and protein design	Chai Discovery is the most defensible leader. Earendil Labs is the biggest new-money bet. Profluent is the most ambitious protein/programming play. EvolutionaryScale, Cradle, and Xaira remain important but less directly proven in drug discovery.
AI small molecules	Insilico Medicine and Recursion remain the leaders. Genesis Therapeutics is the freshest partner-validated private name. Iambic is the clinical-stage private challenger. CHARM, Terray, and Chemify each bring narrower but real strengths.
Self-driving lab / AI lab loop	Recursion has the mature loop. Medra has the freshest autonomous-lab story. Chemify owns chemistry automation. Terray remains credible in small-molecule data loops. Cradle brings feedback loops to protein engineering.
Clinical development AI	Formation Bio leads because it attacks development bottlenecks directly. Owkin is strongest in AI agents and pharma workflows. Recursion and Insilico bring clinical-pipeline evidence. insitro remains strong in disease biology and target selection.
Hot but still proof-discounted	Xaira, Earendil Labs, EvolutionaryScale, and Chai Discovery all matter, but their public evidence is not yet as strong as their ambition, funding, or model story. Chai has the strongest validation in this group.
Older leaders still relevant	Recursion, Insilico Medicine, insitro, Owkin, and Genesis Therapeutics still matter because they kept producing fresh evidence. Older generic virtual-screening stories feel less central today.
Overall top tier	Isomorphic Labs, Insilico Medicine, Recursion, Chai Discovery, and Genesis Therapeutics stand out most consistently after aggregating capital, pharma validation, clinical proof, model ambition, and platform depth.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.