Technology

CuspAI Raises $450 Million to Build an AI Search Engine for New Materials

A Cambridge-based startup backed by Jeff Bezos, Kleiner Perkins, and Nvidia is using generative AI to discover materials that could replace scarce metals and transform manufacturing.

Every modern technology — from smartphone batteries to jet engines to solar panels — depends on materials that are increasingly difficult to source. CuspAI, a British AI startup, has raised $450 million to build the answer: an AI platform that can design new materials on demand, potentially replacing scarce metals like iridium and ruthenium.

The Series B round, announced in July 2026, values the Cambridge-based company at $2.6 billion. Kleiner Perkins and New Enterprise Associates led the financing, with participation from the UK government, Jeff Bezos' investment fund, AMD Ventures, Lux Capital, and other institutional investors. The round comes just one month after reports that CuspAI was seeking $400 million — a sign of the intense investor appetite for AI-driven scientific discovery.

How It Works

Unlike traditional materials science, which relies on trial-and-error experiments that can take years, CuspAI combines generative AI, molecular simulation, automated synthesis planning, and experimental validation in a single pipeline. A researcher describes the desired properties — conductivity, strength, weight, thermal stability — and the AI generates candidate materials. The best candidates are then synthesized and tested in the lab.

This approach is already producing results. CuspAI has launched the AI Materials Foundry, a coalition of more than 45 partners including Nvidia, Meta, and Hyundai. The consortium shares data, computational resources, and validation facilities to accelerate the discovery cycle. The company is targeting materials for semiconductors, advanced batteries, hydrogen production, carbon capture, and alternatives to critical minerals that are concentrated in a handful of countries.

Three Key Facts

$450 million Series B: One of the largest AI funding rounds in Europe, reflecting growing investor confidence in AI-driven scientific discovery.
Targeting scarce metals: CuspAI's first priority is finding alternatives to iridium and ruthenium — metals essential for green hydrogen production and electronics but extremely rare and geopolitically concentrated.
45+ industry partners: The AI Materials Foundry includes Nvidia, Meta, and Hyundai, creating a shared infrastructure for materials discovery that no single company could build alone.

The UK government's participation in the round is notable. It signals a strategic commitment to positioning Britain as a leader in AI-driven science, competing with the US and China in the race to commercialize deep-tech research. The investment follows the UK's broader push to build sovereign AI capabilities, including the £100 million AI Safety Institute and expanded compute access for researchers.

CuspAI's approach represents a broader shift in how scientific discovery happens. Instead of serendipitous lab findings or slow iterative testing, AI systems can now explore vast chemical spaces — millions of potential compounds — and predict which ones will work before a single experiment is run. This "in silico" discovery model could compress decades of materials research into months.

The implications extend beyond materials science. If CuspAI succeeds, AI-driven discovery could become a template for other scientific fields — drug development, catalyst design, battery chemistry, and even quantum computing materials. The startup's $450 million bet is a bet that AI will change not just how we compute, but how we invent the physical world.