Jeremy Wohlwend
Jeremy Wohlwend is a machine-learning researcher and solo developer/creator whose work focuses on open, AI-driven molecular design. He is co-founder and CTO of Boltz, where he helps build molecular-structure, binding-affinity, and generative-design models and tools intended to make advanced biomolecular modeling accessible to scientists.
At a Glance
- Academic laboratories
- Biotechnology companies
- Pharmaceutical companies
- Drug-discovery and preclinical R&D teams
- +1 more
AI Tools by Jeremy Wohlwend
(1)
Boltz
Open Source Biomolecular Structure Predictor
Discussions
No discussions yet
Be the first to start a discussion about Jeremy Wohlwend
Latest News
Boltz expanded its biomolecular AI portfolio with BoltzMol-1 and BoltzProt-1 and launched the Boltz API.
Boltz launched as a public-benefit corporation, introduced Boltz Lab and design agents, announced a $28M seed round, and announced a Pfizer partnership.
BoltzGen was announced as an open-source generative model for designing binders to diverse biomolecular targets.
Boltz-2 was released as an open-source model for joint structure and binding-affinity prediction.
Products & Services
Open-source all-atom foundation model for predicting 3D structures of biomolecular complexes, including proteins, nucleic acids, and small molecules; released under the MIT license.
Open-source biomolecular foundation model that jointly predicts molecular structure and protein-ligand binding affinity; includes controllability features, GPU optimizations, and support for large-scale virtual screening.
Open-source all-atom generative diffusion model and end-to-end pipeline for designing protein and peptide binders against proteins, nucleic acids, and small molecules.
Beta hosted platform providing access to Boltz models and agents with scalable compute, collaborative interfaces, and workflows for small-molecule discovery and protein design.
Market Position
Wohlwend's Boltz work positions open-source biomolecular foundation models as an accessible alternative to closed systems such as AlphaFold3 and as a faster, lower-cost complement or alternative to physics-based free-energy perturbation workflows. Boltz differentiates through permissive licensing, unified structure/affinity modeling, generative binder design, experimental validation, and a hosted product layer for teams that need scalable infrastructure rather than raw model code alone.
Leadership
Founders
Jeremy Wohlwend
PhD in machine learning from MIT (2020–2025), advised by Regina Barzilay; MIT MEng in computer science and electrical engineering (graduated 2017); previously spent about three years as a staff research engineer at ASAPP working on natural-language processing and model compression, and earlier worked in Ed Boyden's Synthetic Neurobiology group on expansion microscopy and connectomics.
Founding Story
Wohlwend's main venture work began at MIT with Gabriele Corso and Saro Passaro in the Regina Barzilay group. The team created Boltz to make powerful biomolecular models broadly accessible through open science, after concluding that closed models and academic-only code releases limited the impact of AI for drug discovery. In January 2026, the project was incorporated as Boltz PBC to build durable products and infrastructure in addition to releasing research models.
Business Model
Revenue Model
Boltz's venture plans to monetize hosted software and infrastructure through usage-based charges for predictions, design campaigns, and large-scale screening, alongside paid enterprise services such as fine-tuned models, customized agents, private-cloud or single-tenant deployments, and production support. The open-source models remain available for academic and commercial use.
Pricing Tiers
Boltz says academics receive free credits upon signing up; larger research grants were planned.
Customers pay for what they run, such as predictions, design campaigns, or large-scale screening, rather than a one-size-fits-all contract.
Target Markets
- Academic laboratories
- Biotechnology companies
- Pharmaceutical companies
- Drug-discovery and preclinical R&D teams
- Scientists and developers building molecular-design agents and pipelines
- Protein, peptide, antibody, and nanobody binder design
- Small-molecule hit discovery and virtual screening
- Biomolecular structure prediction and interaction modeling
- Binding-affinity estimation and prioritization of compounds
- AI-assisted drug discovery and preclinical research workflows
- Academic research and commercial biotech/pharmaceutical model development
- Pfizer
- Teams at all 20 largest pharmaceutical companies, according to Boltz's launch materials
- More than 100,000 scientists across academia and industry, according to Boltz's 2026 launch materials