AI
Boltz raises $110M as pharma giants adopt AI molecular design
London startup reports 50-fold usage surge after launching proprietary models for small-molecule and protein drug design.
Key takeaways
- Boltz raised $110 million Series A from Bessemer Venture Partners to expand its AI platform for molecular drug design.
- The company reports 50-fold usage growth since launching proprietary models BoltzMol-1 and BoltzProt-1 in June.
- Six of the world's 20 largest pharma companies, including Pfizer and GSK, have signed partnerships with Boltz.
- Boltz sells computational capacity through molecular tokens rather than charging per drug discovery program.
- Total disclosed funding reaches $138 million less than a year after the company's founding by MIT researchers.
London-based Boltz has closed a $110 million Series A led by Bessemer Venture Partners to expand its AI platform for computational drug design, the company announced this week. The financing comes as Boltz reports a 50-fold increase in usage volume since launching proprietary molecular generation models in June.
Unlike AI drug discovery startups building their own therapeutic pipelines, Boltz positions itself as infrastructure — selling computational capacity to pharmaceutical and biotech researchers who want to explore potential drug candidates before committing to costly laboratory work.
The company has signed partnerships with six of the world's 20 largest pharmaceutical companies, including Pfizer, GSK, Takeda, and Bristol Myers Squibb, according to European Biotechnology, which first reported the funding details.
From open-source to commercial models
Founded in 2025 by MIT researchers Gabriele Corso (CEO), Jeremy Wohlwend (CTO), and Saro Passaro (research lead), Boltz initially gained traction through open-source AI models that predict molecular structures and binding interactions. The company's Boltz-1 and Boltz-2 models attracted more than 100,000 scientists.
In June, Boltz launched two proprietary commercial models:
BoltzMol-1 designs small molecules that could bind to specific biological targets. In experiments, the model identified confirmed hits against six of 10 challenging targets by testing just 28 to 51 compounds per target.
BoltzProt-1 generates protein designs, including antibody-like binders, for selected targets.
Both models are available through Boltz Lab, a platform for running computational experiments, and via API for integration into existing research workflows.
Usage-based pricing model
Rather than charging per drug discovery program, Boltz sells computational capacity through "molecular tokens" — units that measure the molecular information processed by its AI models. This allows customers to pay according to usage volume. The company reported that molecular token volume for paying customers increased 50-fold following the June launch of its proprietary models.
Why it matters
Boltz's rapid adoption signals a shift in how pharmaceutical companies approach early-stage drug discovery. By positioning AI molecular design as routine computational infrastructure — comparable to sequencing analysis or computational chemistry — Boltz is betting that in silico screening will become standard practice before expensive wet-lab experiments. The 50-fold usage surge suggests major pharma companies are moving beyond pilot projects to production-scale deployment of generative AI for molecular design.
Funding context
Bessemer Venture Partners led the round, with participation from existing investors Andreessen Horowitz, Amplify Partners, Zetta Venture Partners, and Obvious Ventures. New backers include Greycroft, F-Prime Capital, Hummingbird, and Benchling.
The financing follows a $28 million seed round announced in January, bringing Boltz's total disclosed funding to $138 million in under a year. The company's team has grown to approximately 40 researchers and engineers.
Boltz's announcement came one day after London-based Isomorphic Labs raised a $5 billion Series C, one of the largest private biotech funding rounds on record. While the two companies pursue different business models, the back-to-back announcements underscore investor appetite for AI-powered drug discovery platforms.
Boltz plans to invest the new capital in computing infrastructure, data acquisition, model development, and recruitment. The company's stated long-term goal is enabling scientists to progress from therapeutic hypothesis to a molecule ready for human testing entirely through computational design — a challenge that remains scientifically formidable.
Details were first reported by European Biotechnology.