Lila Sciences funding: Every round, investor, and valuation (2026)

Lila Sciences is trying to automate the scientific method. Its AI models propose hypotheses, design experiments, and send them to robotic labs that run the work and feed the results back to the models. The company calls these labs AI Science Factories and its goal "scientific superintelligence," a system that makes discoveries faster than human teams can in biology, chemistry, and materials science.

Lila was built inside Flagship Pioneering, the venture firm that created Moderna, and spent two years in stealth before launching in March 2025. Since then it has raised $550 million across three funding events, all equity: a seed round and a Series A that closed in two parts.

The investor list mixes biotech money with chip makers, sovereign wealth, and pension funds, including Nvidia, Analog Devices, and an Abu Dhabi Investment Authority subsidiary. In June 2026, Bloomberg reported that Lila was in talks for a far larger round. Let's dive into Lila Sciences' funding history.

Lila Sciences went from stealth to a billion-dollar valuation in about six months. Flagship funded the early work, then outside investors took over the larger checks once the labs had results to show.

Flagship Pioneering does not wait for founders to pitch it. It builds companies in its own labs, staffs them with its partners, and funds them until they are ready to raise outside money. Lila came out of that process in 2023, when Flagship merged two internal projects into one company.

Flagship general partner Geoffrey von Maltzahn became CEO, and Flagship founder Noubar Afeyan became co-founder and chairman. The team that followed was built to cover both halves of the idea: geneticist George Church as chief scientist, AI researcher Kenneth Stanley, known for his work on open-ended algorithms, and MIT materials scientist Rafael Gómez-Bombarelli. Flagship has not said how much it spent on Lila before the launch.

Lila came out of stealth in March 2025 with $200 million in committed seed capital, which Flagship led. General Catalyst, March Capital, and Cathie Wood's ARK Venture Fund joined, along with the State of Michigan Retirement System and a wholly owned subsidiary of the Abu Dhabi Investment Authority.

The pitch at launch was a set of early results rather than a product. Lila said its autonomous labs had already produced new antibodies, genetic medicine designs, and materials for carbon capture. None of those claims had been peer reviewed at the time, and the company did not name customers.

Braidwell and Collective Global co-led the first close of the Series A in September 2025. Every major seed investor returned, and new names included Australian pension fund NGS Super and The Mathers Foundation. Lila said the money would scale its AI Science Factories and open new sites in Boston, San Francisco, and London.

The second close came a month later and changed the shape of the cap table. Nvidia's venture arm, NVentures, joined Analog Devices, a maker of sensors and signal chips, and IQT, the nonprofit investor that backs technology for US intelligence and national security agencies. Lila said the $350 million Series A brought its total raised to $550 million.

In June 2026, Bloomberg reported that Lila was in talks to raise about $2 billion, with CalPERS, the largest US public pension fund, and NVentures as anchors. Lila and CalPERS declined to comment. Private-market trackers have since listed a Series B as closed, but Lila has not announced one, and its terms remain unconfirmed.

If the round closes at the reported size, it would be almost four times all the money Lila has raised so far. This article counts only the $550 million the company has confirmed.

Lila's confirmed valuation moved only slightly between the two closes of its Series A. The reported Series B talks would reprice the company in a different league.

Lila did not disclose a valuation for its seed round. Bloomberg's figure for the first close of the Series A does not say whether it is pre-money or post-money, and Reuters described the extension as lifting Lila to more than $1.3 billion. At that price, investors valued Lila at about 2.4 times the capital it had raised.

The Series B talks would mark a step change. A pre-money valuation of about $8.5 billion is roughly 6.5 times the October 2025 figure, reached in about eight months and without a disclosed revenue number. If the full $2 billion comes in, the post-money valuation would be about $10.5 billion.

The jump puts Lila closer to AI model companies than to biotech startups. Its closest rival, Periodic Labs, raised a $300 million seed in September 2025 and was reported in May 2026 to be in talks at about $7.5 billion. Robotics companies building physical AI have seen similar repricing, as Physical Intelligence's funding rounds show.

Lila's investors fall into three groups: the Flagship network that created it, financial investors that led the Series A, and strategic backers that make the chips and sensors its labs run on.

Flagship led the seed round and returned in the Series A. It is the founder, the incubator, and, in all likelihood, the largest shareholder, though its stake is not public. Lila's CEO and chairman both come from Flagship's partnership.

The arrangement is the firm's standard model. Flagship forms a company around a scientific idea, owns a large share from day one, and brings in outside investors only once the idea has early proof. Moderna followed the same path.

General Catalyst and March Capital joined the seed and came back for both closes of the Series A. General Catalyst has a long record in health and AI, and it co-created another AI healthcare company, as Hippocratic AI's funding history shows.

ARK Venture Fund, Cathie Wood's fund for private companies, has also invested in both rounds. Its stake gives retail investors in the fund a rare way into a Flagship company before any IPO.

Braidwell, an investment firm focused on life sciences, and Collective Global co-led the first close of the Series A. They were the first outside firms to lead a Lila round, which moved the company's pricing out of Flagship's hands.

The October 2025 extension brought in investors that also shape Lila's technology. Nvidia sells the GPUs that train AI models, and Nvidia's chip-selling business model benefits from every new class of AI customer. Its venture arm has backed a wide range of physical AI startups, including Skild AI's robot-brain rounds.

In June 2026, Lila announced a collaboration with Nvidia on its BioNeMo Agent Toolkit and Nemotron models. Lila has not disclosed any compute deal or financial terms tied to the investment.

Analog Devices makes the sensors and converters that let lab instruments measure the physical world, which matters to a company running thousands of automated experiments. IQT gives Lila a link to US national security work, a path similar to Palantir's early In-Q-Tel backing.

Lila's cap table includes an unusual amount of long-term institutional capital. A subsidiary of the Abu Dhabi Investment Authority invested in the seed and the Series A, alongside the State of Michigan Retirement System and NGS Super. If the Series B closes as reported, CalPERS would join them as an anchor.

Lila's main cost is physical. Most AI companies spend their capital on compute and talent, while Lila also has to buy, house, and run the lab equipment its models control.

The largest commitment is lab space. In October 2025, Lila leased more than 235,000 square feet at Alewife Park in Cambridge, the largest lab lease in Greater Boston that quarter, and it is moving its headquarters there in 2026. Each AI Science Factory packs automated instruments for biology, chemistry, and materials work under the control of Lila's software.

The company says its goal is to put more instruments under AI control in each lab and to push the system to "superhuman" performance in each field. That requires a mix of robotics, sensors, and lab staff that a software-only AI company does not carry.

Lila trains its own scientific reasoning models and uses Nvidia hardware and software to run them. It has not disclosed its compute spending. The bill is far smaller than the frontier labs carry, as Anthropic's compute-heavy funding rounds show, but Lila must pay for both the models and the experiments that generate their training data.

That data is the core of the strategy. Public scientific data is limited and often hard to reproduce, so Lila produces its own by running experiments at scale and feeding the results back into its models.

The Series A was earmarked for hiring across AI, software, robotics, biology, chemistry, materials, and operations. Beyond Cambridge, Lila has sites in Boston, San Francisco, and London. The company has not published a headcount, and outside estimates range from about 250 to more than 450 employees.

Lila has said it will make money in two ways. Outside R&D teams can pay to use its platform and labs, and the company runs partnered discovery programs that produce new products and intellectual property it can license or spin out. It welcomed its first cohort of customers in October 2025 but has not named them or disclosed revenue.

The early results span several industries. In September 2026, Lila said its AI-run lab screened 2,942 catalysts for green hydrogen production and found six families of palladium-based catalysts that perform as well as ruthenium-based ones, about 17 times faster than a standard lab. In July 2026, it won three Phase I awards under the Department of Energy's Genesis Mission, on projects that include carbon conversion and semiconductor materials.

Lila's growth push is aimed at industries outside biotech. Its site lists therapeutics, chemicals, advanced materials, energy, and aerospace and defense, and the Department of Energy work and the IQT investment point toward government customers. The company is opening its platform to outside teams while it pursues its own discoveries. That tests whether one generalist system can compete with specialists that focus on a single field, such as drug development, where Unlearn's digital twin funding backs AI built only for clinical trials.

The product is moving toward more autonomy in each lab and new models built for scientific reasoning. A Series B of the reported size would pay for more AI Science Factories and the compute to run them, and it would give Lila room to fund its own drug, materials, and energy programs rather than rely only on partners. The company has no public IPO plans.

The main risk is proof. Lila has not yet shown a discovery that earns significant revenue, and the labs carry high fixed costs whether or not customers fill them. Periodic Labs is chasing the same idea with comparable funding, and both compete with the largest AI labs for researchers. Whether the reported Series B closes, and at what price, will show how much patience investors have for a business that sells discoveries rather than software.

Lila Sciences has raised $550 million, according to the company. That includes a $200 million seed round in March 2025 and a $350 million Series A that closed in two parts in September and October 2025. In June 2026, Bloomberg reported that Lila was in talks to raise about $2 billion more, but the company has not announced that round.

Lila Sciences was created by Flagship Pioneering, which led its seed round. Braidwell and Collective Global co-led the Series A. Other investors include General Catalyst, March Capital, ARK Venture Fund, NVentures (Nvidia), Analog Devices, IQT, Altitude Life Science Ventures, Modi Ventures, NGS Super, the State of Michigan Retirement System, and a subsidiary of the Abu Dhabi Investment Authority.

Lila Sciences was valued at more than $1.3 billion after its Series A extension in October 2025, according to Reuters. In June 2026, Bloomberg reported that it was in talks to raise about $2 billion at a valuation of about $8.5 billion before the new money. The company has not confirmed those terms.

No. Lila Sciences is private and has not filed for an IPO. Its shares are not traded on any stock exchange.

Lila Sciences is owned by its founders, employees, and investors. Flagship Pioneering created the company and is likely its largest shareholder. Other major holders include the investors that led or joined its seed and Series A rounds, such as Braidwell, Collective Global, General Catalyst, and March Capital, though exact stakes are not public.