Back to Explore
NewsVibeFix editorialUpdated Nov 18, 2026

Isomorphic Labs Raises $5 Billion, Biotech's Biggest Round Ever — Before a Single Drug Has Entered a Human

On 17 November 2026 Isomorphic Labs raised $5 billion in Series C funding, the largest venture round in biotech history. Greenoaks and Sequoia co-led, Alphabet invested again, and every existing backer joined — yet none of the company's AI-designed drugs has entered a human.

A scientist in a white coat, goggles and blue gloves pipetting green liquid into a test tube in a laboratory

The short version: $5 billion, and no drug has entered a human yet

On 17 November 2026, Isomorphic Labs announced a $5 billion Series C financing round. It is the largest venture round in biotech history, passing the $3 billion record set by Altos Labs at its launch in 2022. The round was co-led by Greenoaks and Sequoia Capital, with Andreessen Horowitz and several other new investors joining, Alphabet investing significantly again, and every existing investor taking part. The company said the round was oversubscribed, but it did not disclose a valuation — Bloomberg reported on 4 November that the company had been negotiating a raise at a valuation of roughly $50 billion, a figure that remains unconfirmed by the company.

The timing is what matters most: the money arrived while not a single drug designed by Isomorphic Labs has yet entered a human being. The company's pipeline is still entirely preclinical, and the official release describes its clinical plans only as "setting the stage for clinical entry", with no date for a first human trial. At the Davos forum in January, Hassabis said he hoped to start first clinical trials before the end of 2026, but the November financing announcement did not repeat that timetable. In other words, $5 billion is not buying a proven drug. It is buying a machine for designing drugs, and investors' trust that the machine will eventually produce one.

Who put up the money

The official release lists a long roster: co-leads Greenoaks and Sequoia Capital, new investors including Andreessen Horowitz, a "continued significant investment" from Alphabet, and all existing investors — Thrive Capital, MGX, GV, CapitalG, Temasek and the UK Sovereign AI Fund. Every existing investor participating is unusual in biotech financing: it usually means insiders judge the progress more optimistically than the public record shows — although it can also mean the round has grown so large that only the existing shareholder circle can absorb it.

The round came only about six months after the last one. In May, the company closed a $2.1 billion Series B; before that, its first external round, in March 2025, was $600 million. According to BioSpace, Isomorphic Labs has raised more than $7.7 billion in total since March 2025. The three rounds were $0.6 billion, $2.1 billion and $5.0 billion — each more than double the one before. That acceleration is itself a signal: the company is spending more because it is shifting from "doing research" to "building capacity". The release says the money will fund "global AI-led bioengineering infrastructure" and "scalable AI-first translational and clinical development infrastructure". In plain language: earlier money mostly paid for scientists and models; this money pays for laboratories, clinical pipelines and getting molecules to patients. Notably, the release also lists a "partner ecosystem" as a use of funds: deepening the understanding of disease biology depends not only on the company's own models but on a network of partnerships with drug companies and academic institutions. A platform company's moat is never just the algorithm; it is the data, relationships and experimental capacity that grow around it, and those cost real money to build.

What IsoDDE actually is

Isomorphic Labs' core asset is called IsoDDE, its proprietary drug design engine. According to the company, it is a suite of foundational simulation models that work natively across multiple therapeutic areas and drug modalities, combined with proprietary agents that navigate and reason over highly complex biological systems to explore vast chemical spaces and design, refine and deliver "exquisitely engineered molecules with optimal drug-like properties". Every word of that is grand, but underneath it sits the oldest problem in pharma: a new drug takes on average more than a decade and well over a billion dollars from target to market, with a failure rate above 90%. Can computation compress that process by an order of magnitude?

The logic runs like this: simulation models first predict how a molecule and a protein interact, with accuracy approaching experimental levels; agents then reason and plan on top of those predictions, deciding which molecule to synthesise and test next; wet-lab results are fed back into the models, closing the loop. President Max Jaderberg said in the release that the team is "making repeatable scientific advancements every week" that "translate directly into better molecular profiles". The key word is "repeatable", not a single spectacular case. Drug discovery's greatest fear is the one-off: a molecule that looks perfect in simulation and collapses in the test tube. What Isomorphic wants to prove is that its engine can deliver good molecules steadily, again and again.

So far the evidence stops at the preclinical stage. The release says internal programmes in oncology and immunology, and partnered programmes, have produced "positive preclinical results". Oncology and immunology are the company's two lead areas — and the two most competitive areas in all of AI drug discovery. Positive preclinical results are a necessary condition, but far from a sufficient one: a tumour shrinking in a mouse and a drug proving safe and effective in humans are separated by the cruellest attrition rate in the industry.

From AlphaFold to drug discovery: five years on the road

Isomorphic Labs was founded in 2021 as a spin-out from Google DeepMind, led by founder and CEO Demis Hassabis. Its scientific lineage runs back to AlphaFold, the protein-structure prediction system that won Hassabis and John Jumper the 2024 Nobel Prize in Chemistry. AlphaFold answered the question "what does a protein look like?"; Isomorphic wants to answer the next one: knowing a protein's shape, can you design a drug molecule that fits into it precisely?

Over five years, the company has taken a different path from a typical biotech. A typical biotech tells its story through one or two lead programmes; Isomorphic tells a platform story: build the engine first, then let the engine generate drug candidates in bulk, with internal and partnered pipelines running in parallel. Partners include Eli Lilly, Novartis and Johnson & Johnson, according to Bloomberg's reporting — the official release does not name them individually. The platform model's upside is a large imagination; its downside is a long validation cycle. Whether the engine is any good can only be judged once the first molecules complete clinical trials, and that takes years.

In August, Hassabis stepped down as CEO of Google DeepMind to become its chairman and Alphabet's chief scientist, putting his full working attention on Isomorphic Labs. That is the other context for this round: the founder is no longer a manager splitting his time between two organisations, but a full-time chief. Three and a half months later, $5 billion arrived. Investors are betting not only on the IsoDDE engine, but on Hassabis himself — a Nobel laureate who has already shown inside Alphabet that he can turn research into products.

Isomorphic Labs funding rounds compared: $0.6 billion in March 2025, $2.1 billion in May 2026, $5.0 billion in November 2026

The funding ladder: each round more than double the last

Put the three rounds side by side and the shape of the curve says more than any single number. In March 2025, $600 million told the story "we have an engine"; in May 2026, $2.1 billion told the story "the engine is producing candidates"; in November 2026, $5 billion tells the story "we are taking candidates into the clinic and building the infrastructure to do it". Each narrative is closer to patients than the last — and burns more cash. For scale: the $5 billion beats Altos Labs' $3 billion record from 2022, and it is more than a fifth of the $23.2 billion that AI drug-discovery startups raised in all of 2026, a record year according to Dealroom.

But a financing record is not a scientific record. Altos Labs took $3 billion and still has no drug on the market; Isomorphic has taken $5 billion and faces the same question. Money buys time, talent and experimental capacity; it cannot buy biology's cooperation. However accurate a drug-design engine's predictions, the human body is the final judge. That investors are willing to pay $5 billion for this story shows they believe AI drug design has moved past the question of "does it work?" into a race over "who gets there first". Who gets there first is still unanswered.

A $50 billion valuation? The number nobody announced

The most speculated-about number in this round is precisely the one the company did not announce: its valuation. Bloomberg reported on 4 November that the company was negotiating at roughly $50 billion; Isomorphic declined to comment, and the announcement said nothing about it. What would $50 billion mean? It is higher than the market value of many mid-sized drug companies that already sell approved medicines, while Isomorphic's revenue comes almost entirely from milestone payments from partners — figures that, as company profiles note, are potential payments contingent on future progress, not cash already received.

Not announcing a valuation is itself a statement. Announcing a sky-high number invites accusations of a bubble; staying silent lets the market guess, with Bloomberg's report as the anchor. Whatever the real figure, one thing is certain: after this round, Isomorphic is among the most highly valued private biotech companies in the world, and it still has no drug in humans. That combination — high valuation, zero clinical products — is rare in biotech. Usually only platform companies command such pricing, and platform valuations ultimately have to be redeemed by clinical data. Hassabis said in the release that the founding conviction was that "biology could be understood and modelled at the most fundamental molecular level and radically accelerated through AI". In 2021 that was a vision; in 2026 it is still a vision, only now it has $5 billion behind it.

Why investors are willing to bet

$5 billion is not one institution's impulse decision. Greenoaks founder Neil Mehta said in the release that discovering new medicines is "an optimization challenge of almost infinite scale", and that Isomorphic is "systematically mapping the path to new medicines". Sequoia's Konstantine Buhler said using AI to improve human health is "one of the most important challenges of our generation". Those sound like standard financing-release lines, but the arithmetic behind them is concrete: returns on traditional pharma R&D have fallen for years, big drug companies find it ever more expensive and slower to develop drugs in-house, so they buy innovation from outside — acquiring biotechs or partnering with platforms like Isomorphic, as Eli Lilly and Novartis have done. For an investor, backing Isomorphic is a bet on several directions at once: the internal pipeline, the partnered pipeline and licensing of the engine itself. Any one of them working out pays off.

There is also a structural factor: Alphabet. Isomorphic is an Alphabet subsidiary, and Alphabet is not only a shareholder but its talent pool and compute backstop. Hassabis is also Alphabet's chief scientist, which means Isomorphic can draw on research resources across Alphabet without building everything from scratch. That structure — incubated by a giant, financed independently — has almost no precedent in biotech: it gives a five-year-old company the focus of a startup and the resources of a giant. It also raises a question: Alphabet holds 75% or more of the company's shares and voting rights, according to UK Companies House records. Investors are betting on Isomorphic, but what they really depend on is Alphabet not pulling back and Hassabis not being called away to fight another fire.

The real test is in the clinic

For Isomorphic, the next eighteen months matter more than the past five years. Once the first human trial starts, the story changes from "how accurate are our models?" to "is our drug safe and effective in people?" Those two questions cannot be answered by simulation; only clinical data can answer them. And the industry record for AI-designed drugs is still blank: to date, no AI-designed drug has ever received regulatory approval. Plenty of AI-designed candidates have entered trials, but none has gone all the way. That blank is both Isomorphic's opportunity and the doubt hanging over it — if the first clinical results for AI-designed drugs disappoint, valuations across the whole sector will take a hit.

There is also a more practical problem: how to spend the money. $5 billion demands an organisational leap — from a research organisation of hundreds of people to a near-pharma company running several clinical programmes at once, managing a global laboratory network and dealing with regulators. Max Jaderberg was named president in November 2025 and took up the role in January 2026, precisely in preparation for that transition. He is a founding member from DeepMind who understands both models and how to organise research. But clinical development is a different craft: trial design, patient recruitment, regulatory communication — each step has its own pitfalls, and AI companies have rarely had the chance to stumble through them.

Partners matter here too. Eli Lilly, Novartis and Johnson & Johnson are both Isomorphic's customers and its validators: their willingness to hand their targets to the company's engine is itself an endorsement. But partnerships have limits — big pharma's most prized assets are always its own pipelines. Isomorphic's internal pipeline will ultimately stand or fall on its own clinical execution, not on its partners' names.

What this money means for AI drug discovery

Zoom out, and the $5 billion Series C is not just one company's financing; it is a pricing event for the whole AI drug-discovery sector. In 2026, AI drug-discovery startups raised a record $23.2 billion, and Isomorphic alone took more than a fifth of it. Investors' judgement is clear: AI drug discovery is no longer a question of "does it work?" but of "who leads it?". AlphaFold's Nobel Prize was two years ago; the market now wants to know whether the people behind AlphaFold can turn protein-structure prediction into real new medicines.

Three things are worth watching next, and none of them is in the financing release. First, when the first human trial is announced — the watershed between "talking about the engine" and "talking about drugs". Hassabis's January mention of "before the end of 2026" is now imminent, and the announcement did not repeat it, which is worth noting. Second, whether big-pharma partners upgrade their collaborations from early exploration to deep partnerships covering the clinical stage; that would be the weightiest outside vote on IsoDDE. Third, which of the internal oncology and immunology programmes produces human data first. Any news on these three will force a re-examination of the $5 billion valuation logic; with no news, all judgements stay on paper.

For ordinary people, the meaning of all this is slower and more concrete. If Isomorphic's engine really can compress the time and cost of developing new drugs, patients will eventually benefit — more drugs for rare and complex diseases may become worth developing. But there is no shortcut: from the first human trial to an approved drug takes five to ten years even if everything goes well. $5 billion buys the right to keep running that long race, not the finish line. The real answer will only arrive on the day the first patient takes a drug designed by Isomorphic. Until then, every number — $5 billion, $7.7 billion, $50 billion — is a bet on the future.

Views 0Comments 0

Comments (0)

ME
0/1000
Loading comments...

Sources

Browse projectsPublish your project

Related articles

A scientist in a white coat and safety glasses examines a sample under a microscope in a bright laboratory, holding a tablet
News
Great at Assembling Molecules, Weak at Picking Drugs: OpenAI's MedChemBench Grades AI's Medicinal Chemistry

On 20 November 2026, OpenAI released MedChemBench: 99 medicinal-chemistry problems from 117 papers and patents, covering 761 molecules. GPT-6.2 Astra scores 84.5% overall — but 95.4% on assembling structures and only 50.4% on choosing compounds. The total flatters; the components tell the truth.

Industry TrendsModel UpdatesTesting & Quality
A developer with glasses types on a laptop late at night at a wooden desk, a large monitor behind showing colourful code, blurred city lights outside the window
News
Run the Middle Layers Three Times: Inkling 2 Mini Puts a 26B Coding Model on Your Laptop

On 20 November 2026, Thinking Machines Lab released Inkling 2 Mini: a 26.4B-parameter looped transformer that runs its middle 12 layers three times per token, fits in about 20GB, and runs offline on a laptop. It leads open-weights models of its size on DeepSWE and SWE-bench Verified — with the caveats printed in the same announcement.

Product LaunchModel UpdatesAI Coding