Boltz PBC, an applied AI research lab, announced a collaboration with Takeda on June 18, 2026 to deploy its most advanced biomolecular AI models across Takeda's research organization. The agreement gives Takeda's discovery teams direct access to BoltzMol-1 and BoltzProt-1, newly released models for structure prediction and generative design, and aims to make AI a practical part of everyday drug discovery work.
The models are built on proprietary biomolecular foundation models and supplement Boltz's open-source tools already in use across the field. The collaboration is part of a broader push to make AI for Science & Research a practical tool for drug discovery teams, with applications ranging from affinity estimation to in-silico molecular design.
How Takeda scientists will access the tools
Takeda scientists will use the Boltz platform through the Boltz Lab interface and the Boltz API, which supports programmatic access and integration into existing workflows. The API is designed to work with large language model agents, letting researchers orchestrate complex design and prediction tasks using natural language.
"We are proud to bring Boltz's most capable models to Takeda's researchers and help make AI a practical part of everyday discovery work," said Gabriele Corso, Co-founder and CEO of Boltz. "From machine learning experts and computational scientists to medicinal chemists and protein engineers, Takeda teams will be able to leverage these models through intuitive user interfaces, APIs and agent integrations. We look forward to seeing how Takeda scientists apply the platform to advance discovery programs."
For researchers looking to build the skills needed to work with these tools, an AI Learning Path for Biochemists provides structured training on molecular design and structure prediction methods.
Fine-tuning and compound ownership
Boltz scientists will also collaborate with Takeda on selected programs to fine-tune models for specific targets. Takeda will retain full ownership of any compounds generated using the Boltz models and platform, protecting the commercial pipeline built from the collaboration.
"At Takeda, we are accelerating drug discovery by integrating cutting-edge science with the latest advances in artificial intelligence," said Hans Bitter, Head of Computational Science & Data Strategy, Takeda Research. "By deploying Boltz's frontier biomolecular models across our research organization, we aim to give our scientists practical tools that can support structure prediction, molecular design and more efficient advancement of high-quality discovery programs."
Why this matters for Science and Research professionals
This deployment signals a concrete shift in how pharmaceutical companies integrate AI into daily research. Scientists who can work with biomolecular models and API-driven workflows will increasingly be sought after. For biochemists and computational researchers, building skills in structure prediction and generative design now translates directly into practical roles on advanced drug discovery teams. The use of models like BoltzMol-1 and BoltzProt-1 across a global research organization provides a template for how AI will be woven into scientific practice, making hands-on expertise with these tools a clear career advantage.
Your membership also unlocks:
