Boltz secures $28m and deepens Pfizer AI collaboration

Boltz has raised $28 million in fresh funding while entering a strategic partnership with Pfizer that underscores the accelerating push by major drugmakers to embed artificial intelligence deeper into biomolecular research. The deal combines Pfizer’s extensive scientific datasets and therapeutic development expertise with Boltz’s open-source foundation models designed to analyse and generate biological data, aiming to shorten discovery timelines and improve the odds of clinical success.

The funding round, disclosed alongside the partnership, strengthens Boltz’s balance sheet at a moment when pharmaceutical companies are reassessing how AI can move beyond pilot projects into core research workflows. Executives involved in the collaboration say the focus will be on applying large-scale models to protein structure prediction, molecular interactions and early-stage target identification, areas that have historically consumed years of laboratory work.

Boltz has positioned itself within a growing cohort of AI-first biotech firms that emphasise open architectures rather than closed, proprietary systems. Its models are trained on publicly available biological datasets and refined through community contributions, a strategy intended to accelerate iteration while maintaining transparency. Under the agreement, Pfizer will contribute anonymised experimental data and domain expertise to help adapt these models for therapeutic use, while retaining control over any drug candidates that advance through its internal pipeline.

For Pfizer, the partnership reflects a broader recalibration of research strategy following a period of uneven productivity across the global pharmaceutical sector. Large drugmakers face rising development costs, higher regulatory scrutiny and a shrinking pool of blockbuster candidates. AI is increasingly viewed as a way to improve target selection and reduce late-stage failures, which account for a significant share of industry losses. Pfizer has signalled that collaborations with specialist technology firms are central to this approach, allowing it to tap external innovation without fully internalising the risk.

Boltz’s leadership has described the funding as validation of its open-source philosophy at a time when many AI life-sciences companies are opting for tightly guarded platforms. Investors backing the round see potential in models that can be audited, adapted and stress-tested by a wider scientific community, particularly as regulators begin to scrutinise algorithmic decision-making in drug discovery. The capital injection is expected to be used to expand computational infrastructure, recruit talent in machine learning and structural biology, and support joint research programmes with pharmaceutical partners.

The collaboration also highlights a shift in how AI vendors and drugmakers share value. Rather than licensing software on a subscription basis, the Boltz–Pfizer arrangement centres on co-development, with AI models evolving alongside real-world laboratory feedback. This iterative loop is designed to improve model reliability while ensuring outputs remain biologically meaningful, a persistent challenge in applying general-purpose AI to complex molecular systems.

Industry analysts note that biomolecular AI has moved from speculative promise to practical application, driven by advances in computing power and the availability of high-quality biological data. Protein-folding breakthroughs earlier in the decade demonstrated the potential impact of machine learning on fundamental biology, prompting a wave of investment into startups seeking to extend those gains into drug design and optimisation. Boltz’s open-source stance differentiates it within this crowded field, but also raises questions about how intellectual property will be protected as models become more capable.

Pfizer’s participation may help address those concerns by providing a clear commercial pathway for discoveries generated through the platform. The company’s global development and manufacturing capabilities offer a route from algorithmic insight to approved medicine, a transition that remains a hurdle for many AI-led biotechs. At the same time, the partnership allows Pfizer to influence the evolution of foundational models without shouldering the full cost of building them from scratch.