CuspAI, a British AI startup focused on materials discovery, is in talks to raise at least $200 million, according to Bloomberg. Although a term sheet has not yet been signed, the deal could value CuspAI at more than $1 billion.
In September 2025, CuspAI raised over $100 million in a Series A round co-led by NEA and Temasek, valuing the company at approximately $520 million. New commercial deals have since increased the valuation to approximately $800 million by this round.
Finding new materials, such as advanced semiconductors, sustainable battery chemistries, and PFAS filtering compounds, typically took more than a decade of trial and error. AI models trained on large material datasets can speed up this process, but these models and datasets often take years to build. CuspAI is currently bringing this technology to market.
CuspAI was founded in Cambridge, UK in 2024 by Dr. Chad Edwards, a chemist and deep tech entrepreneur who helped grow the quantum computing company Quantinuum, and Professor Max Welling, Distinguished Scientist at Microsoft Research, Vice President of Technology at Qualcomm and an AI expert.
CuspAI calls its platform “the search engine of the material world.” Users can input the desired material properties such as strength, conductivity, and thermal resistance, and the system suggests possible chemical compositions up to 10 times faster than traditional methods.
The company develops generative AI models with synthesis in mind, so they can create materials that can be actually manufactured, not just simulated. CuspAI’s partners include Meta, Kemira, and Hyundai Motor Group.
CuspAI works in a rapidly changing field, working with companies like Schrödinger, PhaseCraft, and PASQAL that use physics-based simulation for drug and materials discovery. The company maintains its strategic advantage through proprietary datasets and synthetic model design.
CuspAI plans to use the new funding to grow its platform, hire more staff, and expand its operations in the US and Asia to meet growing customer demand. The company expects its two-part business strategy, including custom materials and in-house development for corporate partners, to accelerate as the platform and its dataset improve.
