CuspAI's core insight is an "inverse design" approach that flips conventional materials science on its head . Instead of starting with a known material and measuring its properties, users start by specifying the exact properties they need—for example, a semiconductor with a specific bandgap, high thermal conductivity, and low reliance on gallium—and the platform generates entirely new molecular and crystal structures designed from scratch to match those specifications .

Series B Funding and Valuation

CuspAI closed a $450 million Series B on July 20, 2026, at a post-money valuation of $2.6 billion . This was a nearly fivefold increase from its $520 million valuation at the Series A in September 2025 . Total capital raised now exceeds $670 million .

The round was co-led by Kleiner Perkins and NEA, with significant participation from Jeff Bezos's Bezos Expeditions and the UK Sovereign AI Fund (the UK government-backed VC fund) . The round was larger than the $400 million initially reported by the Financial Times in June 2026 .

Full List of Series B Backers

Confirmed investors in the Series B round include :

• Co-leads: Kleiner Perkins, New Enterprise Associates (NEA)
• Participants: Bezos Expeditions, UK Sovereign AI Fund, Glade Brook Capital Partners, Lux Capital, AMD Ventures, StepStone Group, Invest-NL, and John Doerr

Earlier investors from the seed and Series A rounds include Hoxton Ventures (seed lead), Temasek, Prosus, Basis Set Ventures, Giant Ventures, and others .

Advisory Board

CuspAI has assembled one of the most decorated advisory boards in AI and science :

• Geoffrey Hinton – Nobel Laureate and Turing Prize winner, often called the "godfather of AI"
• Yann LeCun – Turing Prize winner, Meta's Chief AI Scientist
• Abhi Talwalkar – AMD board member, former CEO of LSI Corporation (joined at Series B)
• Martin van den Brink – former President and CTO of ASML (joined in 2025)
• Lord (John) Browne – former BP CEO, Chair of the Francis Crick Institute, co-Chair of the UK Government's Council for Science and Technology (joined in 2025)

How the AI-Driven Materials Discovery Platform Works

CuspAI's platform is built on an "inverse design" approach that fundamentally flips conventional materials science .

Traditional method: Start with a known material, test it in a lab, and measure what properties it has. This is slow and expensive—often a decade of R&D per material.

CuspAI's method ("Inverse Design"):

1. Specify the target properties – A customer or researcher inputs the exact properties needed (e.g., "a semiconductor with a specific bandgap, high thermal conductivity, and low reliance on gallium") .
2. Generative AI generates candidates – Proprietary generative models produce entirely new molecular and crystal structures that have never existed, designed from scratch to match those specifications .
3. Simulation & filtering – Physics-based simulations and reinforcement learning rapidly evaluate the candidates, ranking them by predicted performance, stability, and synthesizability .
4. Physical validation – Top candidates are sent to partner labs (via the AI Materials Foundry network) for real-world synthesis and testing, compressing what used to take a decade into months .

The platform launched alongside the AI Materials Foundry, a global network of over 45 organizations—including data providers, labs, compute partners, and scientific experts—to orchestrate the full discovery-to-validation pipeline on one agentic platform . Founding members include NVIDIA (providing compute infrastructure) and Meta's Fundamental AI Research Team (developing the Universal Model for Atoms, or UMA) .

Applications for Clean Energy and Semiconductors

• Semiconductors: CuspAI is working with Nvidia and Meta to discover new chip materials that can replace scarce resources like gallium—99% of which is controlled by China—enabling faster, cooler, and more geopolitically secure chips .
• Clean energy: The platform designs materials for better battery electrodes, solid-state electrolytes, and more efficient catalysts for energy storage and conversion . CuspAI co-founder Max Welling has stated the company is also "building molecules to remove forever chemicals from water" .
• Industrial coatings: Discovery of new polymers and coatings with specific durability, conductivity, or thermal properties on demand .

The company claims it moved a carbon capture material from generative design through synthesis and experimental validation in six months during 2025, illustrating the speed advantage of its platform .

CuspAI's core insight is an "inverse design" approach that flips conventional materials science on its head . Instead of starting with a known material and measuring its properties, users start by specifying the exact properties they need—for example, a semiconductor with a specific bandgap, high thermal conductivity, and low reliance on gallium—and the platform generates entirely new molecular and crystal structures designed from scratch to match those specifications .

Series B Funding and Valuation

CuspAI closed a $450 million Series B on July 20, 2026, at a post-money valuation of $2.6 billion . This was a nearly fivefold increase from its $520 million valuation at the Series A in September 2025 . Total capital raised now exceeds $670 million .

The round was co-led by Kleiner Perkins and NEA, with significant participation from Jeff Bezos's Bezos Expeditions and the UK Sovereign AI Fund (the UK government-backed VC fund) . The round was larger than the $400 million initially reported by the Financial Times in June 2026 .

Full List of Series B Backers

Confirmed investors in the Series B round include :

• Co-leads: Kleiner Perkins, New Enterprise Associates (NEA)
• Participants: Bezos Expeditions, UK Sovereign AI Fund, Glade Brook Capital Partners, Lux Capital, AMD Ventures, StepStone Group, Invest-NL, and John Doerr

Earlier investors from the seed and Series A rounds include Hoxton Ventures (seed lead), Temasek, Prosus, Basis Set Ventures, Giant Ventures, and others .

Advisory Board

CuspAI has assembled one of the most decorated advisory boards in AI and science :

• Geoffrey Hinton – Nobel Laureate and Turing Prize winner, often called the "godfather of AI"
• Yann LeCun – Turing Prize winner, Meta's Chief AI Scientist
• Abhi Talwalkar – AMD board member, former CEO of LSI Corporation (joined at Series B)
• Martin van den Brink – former President and CTO of ASML (joined in 2025)
• Lord (John) Browne – former BP CEO, Chair of the Francis Crick Institute, co-Chair of the UK Government's Council for Science and Technology (joined in 2025)

How the AI-Driven Materials Discovery Platform Works

CuspAI's platform is built on an "inverse design" approach that fundamentally flips conventional materials science .

Traditional method: Start with a known material, test it in a lab, and measure what properties it has. This is slow and expensive—often a decade of R&D per material.

CuspAI's method ("Inverse Design"):

1. Specify the target properties – A customer or researcher inputs the exact properties needed (e.g., "a semiconductor with a specific bandgap, high thermal conductivity, and low reliance on gallium") .
2. Generative AI generates candidates – Proprietary generative models produce entirely new molecular and crystal structures that have never existed, designed from scratch to match those specifications .
3. Simulation & filtering – Physics-based simulations and reinforcement learning rapidly evaluate the candidates, ranking them by predicted performance, stability, and synthesizability .
4. Physical validation – Top candidates are sent to partner labs (via the AI Materials Foundry network) for real-world synthesis and testing, compressing what used to take a decade into months .

The platform launched alongside the AI Materials Foundry, a global network of over 45 organizations—including data providers, labs, compute partners, and scientific experts—to orchestrate the full discovery-to-validation pipeline on one agentic platform . Founding members include NVIDIA (providing compute infrastructure) and Meta's Fundamental AI Research Team (developing the Universal Model for Atoms, or UMA) .

Applications for Clean Energy and Semiconductors

• Semiconductors: CuspAI is working with Nvidia and Meta to discover new chip materials that can replace scarce resources like gallium—99% of which is controlled by China—enabling faster, cooler, and more geopolitically secure chips .
• Clean energy: The platform designs materials for better battery electrodes, solid-state electrolytes, and more efficient catalysts for energy storage and conversion . CuspAI co-founder Max Welling has stated the company is also "building molecules to remove forever chemicals from water" .
• Industrial coatings: Discovery of new polymers and coatings with specific durability, conductivity, or thermal properties on demand .

The company claims it moved a carbon capture material from generative design through synthesis and experimental validation in six months during 2025, illustrating the speed advantage of its platform .