A two-year-old Cambridge startup that uses AI to find and design new physical materials has raised $450 million in a Series B round, valuing the company at $2.6 billion. That is five times its valuation from just ten months ago, when CuspAI raised $100 million at a $520 million valuation. All told, the startup has now raised more than $670 million since it was founded.
The round was co-led by US venture firms Kleiner Perkins and NEA, with what the company describes as ‘significant participation’ from Bezos Expeditions, the family office of Amazon founder Jeff Bezos. As reported by Tech.eu, the UK government-backed Sovereign AI Fund also joined the round, alongside Glade Brook Capital Partners, Lux Capital, AMDVentures, Tru Arrow Partners, and StepStone. Existing investors including Temasek, Northzone, and Giant Ventures also participated.
The speed of this valuation jump is striking, even by AI startup standards. It reflects growing investor conviction that materials science is one of the areas where AI can produce real, measurable results in a short timeframe, rather than speculative long-term payoffs.
CuspAI’s core idea is relatively simple to explain. Scientists and engineers define the properties they need in a material, and CuspAI’s platform searches for and designs candidates that match. Think of it as a search engine for matter. Current customers include ASML, the Dutch company that makes the machines used to print semiconductor chips, and Meta. The industries CuspAI targets include:
- Semiconductor manufacturing
- Clean energy
- Carbon capture
- Water treatment
- General manufacturing
The company counts AI researchers Yann LeCun and Geoffrey Hinton among its advisers, two of the most cited names in the field. LeCun leads AI research at Meta, while Hinton, often called the ‘godfather of deep learning’, won a Nobel Prize in Physics in 2024 for his foundational work on neural networks.
CuspAI plans to use the new funding to expand across the US, Asia-Pacific, and Europe. Central to that plan is what it calls its ‘AI Materials Foundry’, described as a global network of data, labs, computing infrastructure, and scientific knowledge aimed at helping researchers design new materials faster and at lower cost. Founding partners of that network include Nvidia, Meta, Samsung, and Hyundai, along with more than 40 other companies.
John Doerr of Doerr Capital, one of the investors, put the opportunity plainly: ‘Materials discovery has quietly been a bottleneck behind progress in energy, semiconductors, and climate. What’s remarkable about Cusp is that a single platform is already solving hard problems across industries, from stripping forever chemicals from drinking water to designing next-generation chip materials.’
That framing matters for anyone watching the broader AI investment picture. Much of the capital flowing into AI right now is going into foundation models and infrastructure. CuspAI represents a different bet: that AI applied to a specific scientific domain can produce commercial value quickly enough to justify a multi-billion dollar valuation at an early stage. The involvement of the UK’s Sovereign AI Fund also signals that the British government sees materials AI as a strategic priority, not just a niche research area.
