Saturday, November 21, 2026Sat, Nov 21 Updated

Mergers, acquisitions and startup deals, and what they mean.

Boltz Raises $110 Million Series A for AI Drug Discovery Models

Bessemer led a $110 million Series A for Boltz, an AI startup whose models help drug researchers find small molecule hits faster.

Close-up of a scientist using a pipette in a laboratory.
Photo: Leiliane Dutra / Pexels

Boltz PBC, a startup that builds artificial intelligence models for pharmaceutical researchers, has raised $110 million in a Series A round, according to SiliconANGLE. Bessemer led the investment, the company said on Wednesday.

Greycroft, Andreessen Horowitz, Amplify Partners, F-Prime Capital and Hummingbird were among the new and returning backers that joined. The company did not disclose a valuation in the source report.

What Boltz sells

Drug discovery projects often begin by looking for small molecules that bind strongly to a pathogen, a property called binding affinity. Researchers commonly review hundreds of thousands of candidates per project, a process that can take months and cost millions of dollars, according to SiliconANGLE.

Boltz offers a model called BoltzMol-1 that generates binding affinity data by reviewing only a few dozen promising molecules. The output is a set of small molecule hits, which are early clues that show researchers where to focus. Teams can adapt the model to their constraints. A group with a limited number of compounds can restrict it to molecules that can be made from them, and when materials are easy to get, it evaluates a much larger number of candidates.

The company has a second model, BoltzProt-1. It can generate tens of thousands of protein designs based on data from small molecule hits and automatically reviews each design to judge whether it could be used to develop a new drug. According to the company, its evaluation method showed “higher success” than several competing models in internal tests.

Results and customers

Boltz staff asked BoltzMol-1 to look for hits across 10 challenging, therapeutically relevant targets. Per the company, the model found hits for six of them. These are internal results reported by Boltz, not an independent evaluation.

Boltz sells its models through an application programming interface billed by token usage. The company says customers’ token consumption has grown by a factor of 50 since both models launched in June. It counts six of the world’s 20 largest pharmaceutical companies as clients.

Why it matters for investors and the field

The funding gives Boltz money to compete with larger AI labs. SiliconANGLE reported that Boltz will use it to grow its headcount and buy more computing infrastructure, which should help it keep pace with leading labs such as newly public Anthropic PBC and OpenAI Group. Over the past year, both of those companies have introduced models adapted to biology research.

For pharmaceutical customers, the pitch is fewer molecules to test per project and less time spent on each analysis. Binding affinity evaluations do not produce drug discoveries on their own, only the hits that point researchers toward promising areas.

What comes next

Boltz has already started training a new generation of molecule design algorithms. The company says they are “showing substantial improvements” over its current models. No release date was given.

Source: SiliconANGLE