Solid-state batteries keep being five years away. Here's what's actually shipping.

Every technology has its running joke. Fusion is thirty years away and always will be; solid-state batteries are five. I've sat through enough supplier presentations to recite the promise from memory: double the energy density, charging in minutes, no thermal runaway, coming soon. What's changed, as of early 2026, is that "soon" has finally started producing invoices. Just not the ones the headlines implied.

Why everyone wants this battery

A conventional lithium-ion cell moves ions through a flammable liquid electrolyte. Replace that liquid with a solid — a ceramic, a sulphide, a polymer — and several good things follow at once. You can safely pair the electrolyte with a lithium-metal anode, lifting energy density substantially, which means lighter packs or longer range. Fast charging becomes less damaging. And the fire risk that dictates so much of today's pack engineering — the reason your airline is fussy about where your power bank goes — largely evaporates. On paper it is simply a better battery. Paper, as ever, is not a factory.

What is genuinely shipping

Semi-solid EV packs in China. The most visible deployments are "semi-solid-state" cells — a gel-like electrolyte, a pragmatic halfway house. NIO has offered a 150 kWh semi-solid pack (cells from WeLion) enabling four-figure-kilometre range demonstrations, and several Chinese manufacturers now sell semi-solid packs in limited volumes. Purists note, correctly, that these are not true solid-state cells. Customers note, equally correctly, that they exist.

Small devices, quietly. True solid-state cells have shipped for years where the batteries are tiny: thin-film cells in medical implants, sensors and some wearables, and small-format cells from Japanese and Taiwanese makers in industrial kit. Manufacturing a postage stamp is not manufacturing a car boot, but it proves the chemistry in the field.

Automotive samples on test tracks. QuantumScape has been shipping B-sample cells to Volkswagen's battery arm; Factorial's cells have been road-testing with Mercedes-Benz since 2025; Toyota — the most prolific patent-holder in the field — continues to target genuine production around 2027–28, with Samsung SDI and others running pilot lines on similar timelines. B-samples are a real milestone: it means cells representative of production design, being validated in actual vehicles. It is also, honestly, still the rehearsal rather than the show.

Purists note these aren't true solid-state cells. Customers note, equally correctly, that they exist.

The unglamorous bottleneck

The remaining problem is not discovering a material; candidate chemistries abound. It is manufacturing: producing kilometres of defect-free solid electrolyte, keeping fragile ceramic layers intact under the pressure and vibration of a car, managing the interfaces where solid meets solid (liquids are forgiving; solids crack), and doing all of it at a cost per kilowatt-hour that survives comparison with conventional lithium-ion — a moving target that gets roughly 5–10% cheaper every year. The incumbent is not standing still, and "better than today's batteries" means beating tomorrow's price sheet, not today's.

This is why sensible forecasts now stagger adoption: premium vehicles and aviation-adjacent niches first (where cost matters less than weight and safety), volume cars later, and grid-scale storage possibly never — bulk storage cares about cost per cycle, not density, which is why data-centre operators wrestling with the UK's grid squeeze are buying conventional chemistries by the shed-load instead.

What it unlocks beyond cars

The second-order effects interest me more than the range figures. Dense, safe batteries change product design wherever weight and fire risk currently bind: laptops that give a full multi-day charge to power on-device AI workloads, delivery drones with useful payloads, and humanoid robots — whose two-hour runtimes are among the quiet reasons they remain a warehouse story for now. Batteries are the least glamorous dependency of nearly every futuristic product on this site.

How to read the next announcement

A field guide for the press releases you'll see between now and 2030. Check three things. First, semi-solid or true solid-state — both legitimate, very different claims. Second, samples or sales: "shipping B-samples to a partner" and "shipping products to customers" are separated by years and billions. Third, the cost sentence: if an announcement mentions energy density but not cost per kilowatt-hour, the writer knows exactly what they're omitting. As of early 2026, the honest position is that solid-state has finally crossed from perpetual promise to early, narrow reality. The five-years-away joke is retiring. Slowly. In limited volumes. Like everything else in batteries.