Solid-State Batteries Spent a Decade as Vaporware. Seven Companies Just Built Cells.

CALB Chairwoman Liu Jingyu held up a 60 Ah solid-state cell at China's annual Two Sessions conference on March 10. It weighed 490 grams. She claimed 450 Wh/kg. A conventional lithium iron phosphate cell of the same capacity weighs roughly 1.2 kilograms — nearly three times the mass.

The cost premium: 100% over liquid-electrolyte equivalents, roughly 0.85 yuan/Wh ($0.12/Wh). That number matters more than the energy density. At $0.12/Wh, nobody's putting these in a $25,000 sedan.

Samsung SDI doesn't care about sedans either. Its own all-solid-state prototype, unveiled at InterBattery 2026 on March 9, came with a slide deck titled "AI thinks, Battery enables." The target customer: humanoid robots.

The Robot Premium

The press releases bury this under EV range claims, but the first mass market for solid-state batteries almost certainly won't be electric vehicles.

Robots have limited internal volume and generate violent power spikes during movement. They need energy per cubic centimeter that conventional cells can't deliver. More importantly, the battery is a smaller fraction of total system cost in a $50,000 humanoid robot than in a $30,000 car — the manufacturer can absorb a 2× cell premium without pricing the end product out of its market.

CALB's own roadmap confirms the sequencing. Robot and eVTOL deliveries: Q4 2026. Automotive small-batch production: 2027. Samsung SDI's mass-production target for the pouch-type cell: second half of 2027, starting with robotics and "physical AI."

Same sequencing every expensive new chemistry follows: high-margin, low-volume buyers first. Ride the learning curve. Enter the mass market once yield problems are someone else's sunk cost. Goldman Sachs estimates the humanoid robot market at $38 billion by 2035 — a large enough sandbox to fund the manufacturing ramp that EVs demand but can't yet pay for.

Seven Companies, Hardware in Hand

For years, Toyota, Nissan, BMW, Honda, CATL, SES AI, Solid Power, and others cycled through the same press cadence: promising solid-state "within two years," then quietly pushing timelines. Toyota alone has been saying "by 2025" since 2017.

Q1 2026 is different. Seven companies now have physical cells — not investor decks:

Company	Cell Type	Wh/kg	Status (Q1 2026)
CALB	60 Ah sulfide	450	Prototype; robot/eVTOL deliveries Q4 2026
Samsung SDI	Pouch all-solid-state	—	InterBattery 2026 unveil; mass prod. H2 2027
QuantumScape	Ceramic/Li-metal (QSE-5)	~500	Eagle pilot line running; cells shipped to VW/PowerCo
Geely	All-solid-state pack	~400	First pack complete; vehicle validation 2026
Dongfeng	350 Wh/kg solid-state	350	Winter testing done; mass prod. Sept 2026
FAW Hongqi	Tiangong 06 prototype	—	Prototype vehicle shown Jan 2026
Chery	"Rhino S" engineering samples	—	Production facility online; vehicle testing 2026

Five of the seven are Chinese. That's what State Council battery-tech prioritization and national solid-state standards (landing July 2026) buy you. The money tracks: Top Technology is building a 15 GWh factory in Inner Mongolia for 6 billion yuan ($830 million). WELION New Energy has a 15 GWh joint venture in Jiangning. Construction contracts, not research grants.

What's Still Broken

Three production-scale problems separate a demo cell from a factory line:

Interface resistance. Two solids conduct ions worse than a solid bathed in liquid. Every company in that table fights this with hot pressing, surface coatings, or in-situ formation — but nobody has published cycling data from a continuous manufacturing line. Lab cells pressed under controlled conditions are a different animal from cells vibrating inside a robot chassis.

Dendrite growth at high current. Lithium-metal anodes are the whole point, but lithium grows metallic filaments through the electrolyte during fast charging. QuantumScape's ceramic separator blocks them at moderate rates. CALB claims 6C charging. Nobody has shown long-cycle data at 6C from a production-format cell.

Sulfide stability. The highest-performance solid electrolytes — the LGPS family, ionic conductivity up to 12 mS/cm — generate hydrogen sulfide on contact with ambient air. Manufacturing must be inert atmosphere from start to finish. Samsung SDI and QuantumScape sidestep this with oxide-based chemistry but sacrifice conductivity. CALB and several Chinese manufacturers are betting the performance advantage justifies the cleanroom premium.

The Bottom Line

Seven companies with hardware. Pilot lines running. China drafting national standards. The solid-state battery is no longer hypothetical — but it's not a car story yet. The cost premium locks it out of mass-market EVs until cumulative production drives manufacturing yields up, and that takes years. Robots, drones, and military systems will pay the premium today for density conventional cells can't match. If you're waiting for a $30,000 solid-state EV, think 2029. If you're building humanoid robots, the battery you wanted is shipping this year.