Markets

Biotech's Biggest Venture Round Yet Buys an Engine, Not a Drug

Isomorphic Labs raised biotech's largest venture round, $5 billion, before naming a development candidate. What the money buys, and the clinical scoreboard AI drug design still has to clear.

A dawn-lit glass conference room high above a city, a tablet on the table showing an abstract ribbon of light

Isomorphic Labs announced a $5 billion Series C on Nov. 17, co-led by Greenoaks and Sequoia Capital, with Andreessen Horowitz and other new investors, a significant investment from Alphabet, and participation from all existing backers: Thrive Capital, MGX, GV, CapitalG, Temasek and the UK Sovereign AI Fund. By BioCentury's count it is the largest venture round in biotech history, surpassing the $3 billion Altos Labs raised, and it landed before the company named a single development candidate. Isomorphic did not disclose a valuation. Bloomberg reported earlier in November that it was seeking funding at a valuation of as much as $50 billion.

What the money is for

The round arrives six months after raising $2.1 billion in a Series B in May, bringing disclosed raises to roughly $7.7 billion once the $600 million first external round of 2025 is counted. Isomorphic says the money funds IsoDDE, its drug design engine, and the preclinical pipeline built with it, and separately that it is investing in translational and clinical development infrastructure to move programs into human testing. Demis Hassabis, the company's chief executive and a Nobel laureate in chemistry for the AlphaFold work he did with John Jumper, said the round lets Isomorphic build foundational biology models, deepen its understanding of disease and drive its first therapeutic programs toward clinical trials. Preclinical results have been described as positive, though the company has not disclosed detailed preclinical data. The internal pipeline spans multiple modalities and concentrates on oncology and immunology. IsoDDE, the company says, outperforms AlphaFold 3 on a protein-ligand structure prediction benchmark and identifies novel binding sites.

The scoreboard it still has to beat

The clearest clinical test of an AI-designed drug so far comes from a competitor. Insilico Medicine's rentosertib, a TNIK inhibitor for idiopathic pulmonary fibrosis, ran a 12-week randomized, double-blind, placebo-controlled Phase 2a trial in 71 patients, with arms of 30 mg once daily, 30 mg twice daily, 60 mg once daily and placebo (18, 18, 18 and 17 patients). The primary endpoint was the share of patients with at least one treatment-emergent adverse event: 72.2%, 83.3%, 83.3% and 70.6% across those arms, rates the authors describe as similar. Lung function was a secondary endpoint. Mean change in forced vital capacity at the highest dose was +98.4 ml (95% CI 10.9 to 185.9) against -20.3 ml (95% CI -116.1 to 75.6) for placebo, a positive point estimate attached to a confidence interval wide enough to leave real uncertainty about the size of the effect.

The same paper states that AI-discovered drugs have experienced similar levels of Phase 2 failure as drugs found the traditional way, and that none had progressed through Phase 3 when the results were published in June 2025.

Why the checks keep coming

Pharma has already put money behind the thesis. Isomorphic disclosed alliances with Eli Lilly and Novartis in 2024 worth about $3 billion combined including milestones, and added Johnson & Johnson as a partner at the start of 2026. President Max Jaderberg said the company's researchers and drug hunters "are routinely solving complex biological problems that have stalled traditional research for decades." Alphabet spun Isomorphic out of DeepMind in 2021 and made a significant investment in this round. Hassabis stepped back as chief executive of Google DeepMind about three and a half months before the round was announced and kept the Isomorphic role.

Category-wide, the money is moving the same way. Drug discovery startups raised $23.2 billion in 2026, a record for the category according to Dealroom, and the largest AI labs keep entering the field: Anthropic announced a drug discovery program in June, and OpenAI partnered with Novo Nordisk in April.

What would settle it

None of this resolves the question the record round is a wager on, which is whether AI-designed molecules clear the clinical bar at the rates ordinary drugs do. The capital answers a different question: investors, Alphabet and three large drugmakers have concluded the engine is worth building before the molecules prove themselves. The markers that would answer the first question are concrete. A named development candidate would be the first, an investigational new drug filing the second, a Phase 1 readout the third. Until one lands, the money buys a team, an engine and time. The clinical record for AI-designed drugs in humans is still thin, and its most complete readout is a 71-patient Phase 2a trial run by a competitor.