Peak Energy Says Its Sodium-Ion Batteries Can Outlast Lithium. GM Is Funding the Bet.

Peak Energy, a Colorado battery startup, is building a $71 million sodium-ion gigafactory near Sacramento, and General Motors is underwriting the bet. The chemistry has lived in EV labs and Chinese cell lines; the product is grid storage, the niche where U.S. utilities have been waiting for a sodium-ion option.

The roughly 17,000-square-meter facility, announced this month, is slated to come online in 2027 with four gigawatt-hours of annual capacity. By Peak's own projection, that is enough to back up about four million homes for an hour. The site is a sharp departure from the auto industry's sodium-ion story, which has lived in labs and EV prototypes; Peak is selling hours-long discharge for utilities, the category that lithium iron phosphate (LFP) has owned since 2022.

The timing cuts two ways. The U.S. grid is absorbing an unusual load growth, much of it from the data centers training and running AI models. 2025 also saw both Natron Energy and Bedrock Materials shut their U.S. sodium-ion lines. The two failed attempts used Prussian Blue analogs and traded on cheap raw materials rather than grid economics. Peak's choice of sodium iron pyrophosphate (NFPP) cathodes is the same chemistry CATL has adopted at much larger scale.

The falsifiable claim sits in a product called the GS1.1. Peak says the system delivers a 20-year service life, about 20,000 charge-discharge cycles, and 80% capacity retention at the end of it. The standard LFP pack used in U.S. storage today typically delivers 70% retention after 8,000 cycles, by Peak's benchmark, so the durability gap is two-and-a-half-fold. The cells can also run at roughly double the operating temperature of a typical LFP cell without damage, which lets the company strip the active liquid cooling and HVAC loops that eat auxiliary power on every LFP installation. Peak puts the lifetime-cost advantage at roughly 20% over LFP, even though its per-cell price still trails LFP today.

The anchor is GM. The automaker's cell-research arm and GM Ventures have invested in Peak, and the two are co-developing the chemistry and the pack. Three independent power companies, Jupiter Power, Energy Vault, and RWE Americas, have already signed offtake agreements for the Sacramento output.

picture suggests. In April 2026, CATL signed a 60 GWh grid-storage supply deal with HyperStrong, fifteen times Peak's planned first-phase capacity. Benchmark Mineral Intelligence projects sodium-ion will hold less than 1% of new U.S. storage this year and under 4% by 2030; globally, the firm puts the share around 5% by the same horizon.

There is a second supply chain inside the battery story. Peak is still buying commercial cells from Chinese suppliers while its Sacramento line ramps, and most sodium feedstock processing, even for trona deposits under Wyoming's Green River Basin, still routes through Chinese refiners. A U.S. sodium-ion cell made from Chinese active material would not satisfy the domestic-content rules that increasingly drive federal storage incentives.

The signal that the bet is working: the GS1.1 ships on schedule, holds 80% capacity after 20,000 cycles in a third-party lab test, and pulls enough auxiliary-load savings on a Jupiter Power or RWE site to validate the 20% lifetime-cost edge. The signal that it is failing: another year of buying Chinese cells, an LFP price drop that reopens the cost gap, or a GM executive change that walks the partnership back to EVs. Peak says the Sacramento line starts producing in late 2027.