Holyvolt Raises 45 Million Dollars at a 245 Million Valuation to Let AI Pick the Next Battery Chemistry

Holyvolt, a startup that builds AI tools for designing and testing new battery materials, has raised 45 million dollars from investors including Zone II, CC.VC and a group of finance-sector angels, with Fifth Wall described as an existing investor. Fortune reported that the financing was all equity and valued the company at about 245 million dollars, citing an unnamed source. Holyvolt did not confirm those terms, and the announcement did not label the round as a Series A or a Series B.

The company's product is less a model than a loop. Its Discovery Engine chooses which experiments are worth running, an automated laboratory then builds and tests actual battery cells, and the measurements feed back into what the system proposes next. Holyvolt previously combined with Wildcat Discovery Technologies, which supplies the physical testing capability behind that loop, though that combination predates this financing and is not part of the new announcement.

The first commercial target is a specific chemistry family: disordered rocksalt cathodes that use neither nickel nor cobalt. The company says it aims for initial commercialization in 2027, a date it describes as a plan rather than a validated milestone. Both metals have been the focus of supply chain anxiety for automakers, and a cathode that removes them would change the cost and sourcing assumptions behind a large share of planned electric vehicle production.

Battery development has historically been slow because the search space is enormous and physical verification is expensive. A single promising material has to survive synthesis, cycling, thermal and safety testing before anyone can quote a number that a procurement team will accept. The argument for AI here is not that it replaces the laboratory but that it stops the laboratory from spending months on candidates that a model could have ranked lower.

The gap between a promising sample and a manufacturable material is where most materials startups stall. Meaningful comparisons require disclosed test conditions, reproducibility across batches, and an explanation of which trade-offs, such as energy density against cycle life, the design is optimizing. Holyvolt has not published those results, and customer qualification timelines have not been disclosed.

The round is a useful marker of where applied AI money is going. At a 245 million dollar valuation, Holyvolt is priced far below the general-purpose model companies that dominate AI headlines, and its investors are buying a claim about an industrial bottleneck rather than a claim about intelligence. The test over the next 18 months is whether the loop produces a material a battery maker will actually qualify, or only a better-filtered list of candidates.