Boltz Raises $110 Million Series A to Scale AI Drug-Discovery Models

London-based Boltz, an MIT spinout, raised $110 million in a Series A led by Bessemer Venture Partners to expand its open-source AI models for predicting protein-drug binding.

Key Takeaways

  • Boltz raised $110 million in a Series A led by Bessemer Venture Partners, announced November 19, 2026.
  • Total disclosed funding reaches roughly $138 million, including a $28 million seed round earlier in 2026.
  • Boltz builds open-source AI models that predict protein-drug binding, competing with Google DeepMind's closed-source AlphaFold3.
  • The company has pharma partnerships with Pfizer, GSK, Takeda and Bristol Myers Squibb.
  • New funds will go toward compute, data and expanding Boltz's roughly 40-person London research team.

Boltz, a London-based AI drug-discovery startup spun out of MIT, has raised $110 million in a Series A round led by Bessemer Venture Partners, the company confirmed November 19, 2026. The round follows a $28 million seed raised earlier in 2026, putting Boltz's total disclosed funding at roughly $138 million in under a year. Existing backers Andreessen Horowitz, Amplify Partners, Zetta Venture Partners, Obvious Ventures and Nvidia returned, joined by new investors Greycroft, F-Prime Capital, Hummingbird and Benchling.

What does Boltz's technology actually do?

Boltz builds AI models that predict how strongly a potential drug molecule will bind to a target protein, and proposes new molecules designed to hit a specific target. That prediction step is normally one of the slowest and most expensive parts of early drug discovery, requiring either lab assays or expensive computational chemistry. Boltz's models aim to compress that into a fast, software-driven screening step, letting researchers narrow thousands of candidate molecules down to the handful worth testing in a lab.

Unlike Google DeepMind's AlphaFold3, which remains closed-source for commercial use, Boltz has built its models in the open. That choice has let pharmaceutical companies inspect, adapt and run the models on their own infrastructure, which the company credits for a faster path to its current pharma partnerships.

Who is already using it?

Boltz has disclosed collaborations with Pfizer, GSK, Takeda and Bristol Myers Squibb, generating revenue through both platform access and direct partnership work. That is a notably long list of named pharma customers for a company barely a year removed from its MIT origins, and it is the detail investors are pointing to as evidence the open-source approach converts into paying enterprise deals, not just academic citations.

What will the new money fund?

Boltz says the Series A will go toward buying more compute and data and expanding its research team, which currently stands at roughly 40 AI researchers based in London. More compute means the company can train larger, more accurate binding-prediction models and run more molecule-design cycles per customer engagement, which is the main lever available to compete with AlphaFold3 on raw predictive accuracy.

How does Boltz compare to AlphaFold3, and what comes next?

AlphaFold3 remains the dominant closed-source benchmark in protein structure and binding prediction, built by a far larger, more resourced team inside Google DeepMind. Boltz is betting that openness, speed of iteration, and direct pharma relationships can offset DeepMind's scale advantage, similar to how open-source language models have chased closed frontier labs. The next signal to watch is whether Boltz's pharma partners convert their current collaborations into drug candidates that reach preclinical testing, which would be the first hard evidence the platform's predictions hold up outside a benchmark leaderboard.

Category: Funding & Investments

Tags: AI Models Drug Discovery Bessemer Venture Partners Boltz AlphaFold3 protein structure prediction

Related Articles

Frequently Asked Questions

How much did Boltz raise and who led the round?

Boltz raised $110 million in a Series A round led by Bessemer Venture Partners, announced November 19, 2026.

How much has Boltz raised in total?

Approximately $138 million, combining the new $110 million Series A with a $28 million seed round raised earlier in 2026.

What does Boltz's AI do?

It predicts how strongly potential drug molecules bind to target proteins and proposes new molecules designed to hit specific targets, speeding up early-stage drug discovery.

Who are Boltz's customers and competitors?

Boltz has partnerships with Pfizer, GSK, Takeda and Bristol Myers Squibb, and competes most directly with Google DeepMind's closed-source AlphaFold3.