EV Battery Technology Race: Sodium-Ion Ships as Solid-State Lags

The short version

For most of the past decade, talk about life after lithium-ion meant talk about solid-state batteries. By late 2026 the picture has changed. The chemistries actually reaching cars are mostly the less flashy ones. Sodium-ion cells are going into a mass-produced Chinese hatchback. Silicon-carbon anodes are shipping in a Mercedes. General Motors has picked a factory for lithium manganese-rich cells. Solid-state is moving into pilot production but has not reached showrooms at scale, and lithium-sulfur is still mostly flying drones.

One caveat on the framing: two of these five technologies are not really "beyond lithium-ion." Silicon anodes and LMR cathodes are upgrades inside the lithium-ion family. That matters, because the technologies that fit existing factories are the ones arriving first. Here is where each one stands and how much of the coverage holds up.

1. Sodium-ion: the first real break from lithium

The most concrete milestone this year is sodium-ion. CATL and Changan presented what they call the world's first mass-production passenger car with sodium-ion batteries, a version of the Changan Nevo A06, and set a mid-2026 market launch.1420 CATL rates its Naxtra cell at up to 175 Wh/kg, which is roughly the same as the lithium iron phosphate (LFP) cells used in today's cheaper EVs.15 The 45 kWh pack is rated for more than 400 km of range, and CATL expects 500–600 km as the supply chain matures.1316

Sodium's real selling point is cold weather. CATL says the cell keeps more than 90% of its capacity at –40 °C and delivers almost three times the discharge power of a comparable LFP pack at –30 °C.11 Sodium-ion can also use aluminium foil on both electrodes instead of copper on one, which cuts current-collector cost by about 70%, according to industry analysis.19

The rollout is deliberately small. A CATL executive said 10,000 to 20,000 EVs will carry the company's sodium cells this year.18 Capacity plans are much larger. CATL has announced a 5 billion yuan investment for 40 GWh of new sodium capacity in Fujian.15 It has also signed a three-year, 60 GWh sodium supply deal with HyperStrong, which is for grid storage, not cars.17

The reporting disagrees on cost. One outlet says CATL expects sodium cells to be about 30% cheaper than LFP.13 Another notes that CATL has published no standard pack price and did not disclose what the HyperStrong deal is worth.17 Accounts of timing also differ. One says the A06 has been coming off production lines since February.16 Others describe February as the unveiling, with sales starting mid-year.1420 The safest reading is that sodium-ion is now commercial but at small volume. Its cost advantage is still a projection, and it depends on reaching scale.16

2. Silicon anodes: the drop-in upgrade

If sodium is the most notable chemistry change, silicon may matter more for range in the next few years. Silicon can theoretically store about 4,200 mAh/g, compared with 372 mAh/g for graphite. The catch is that it swells dramatically when charging, which has historically destroyed cells.2

That engineering problem now looks largely solved. Group14 began EV-scale production of its SCC55 material in Sangju, South Korea, in March. The plant is sized for up to 2,000 tonnes a year, enough for about 10 GWh of battery capacity.6 Sila raised $300 million in July to expand its Moses Lake, Washington, plant, which started production last September with about 2 GWh of capacity. The expansion is meant to supply more than 100,000 EVs.4 Sila has supply agreements with Mercedes-Benz and Panasonic.4 The electric Mercedes G580 can be ordered with Sila-based cells for up to 20% more range, from 473 km to about 568 km on the WLTP cycle.3

Supply-chain politics are pushing silicon as much as performance is. Chinese companies control about three-quarters of graphite supply, and silicon offers automakers outside China a way around tariffs.4 Group14 says one tonne of its material can replace about five tonnes of graphite.6 GM's battery team has also said publicly that it is "deep on silicon."7

The gap between lab and road shows up here too. Sila says its material can store up to 40% more energy than graphite.4 In practice, the figure Mercedes reports at the vehicle level is around 20%.3 Long-term durability has not been settled. Independent analysts note that real-world fleet data will be the real test.10

3. Lithium manganese-rich: Detroit's answer to LFP

The newest big commitment comes from GM. In late September the company said its Ultium Cells joint venture with LG Energy Solution will upgrade its Spring Hill, Tennessee, plant to mass-produce prismatic LMR cells. Work starts this year and finishes in 2028, and the first vehicles are due the same year.4448 GM says the chemistry has 33% more energy density than LFP at a similar cost.47 It is aimed at full-size trucks and SUVs, with more than 400 miles of EPA range.44 GM officials have previously suggested the switch could save at least $6,000 on a Silverado EV.43

This is a strategic turn for GM. Battery chief Kurt Kelty has said LFP may never "earn its way" into GM's vehicles and has called LMR the company's "workhorse."42 Ford, by contrast, is still committed to LFP while developing its own LMR cells.49 The main technical risks are durability, capacity loss and fast charging.42 LG Energy Solution says it has fixed a key gas-generation problem, producing a cell that keeps 92% of its capacity after 883 cycles.50 No consumer degradation data exists yet.45

4. Solid-state: real progress, overstated timelines

Solid-state still gets the most attention, and the coverage disagrees on it more than on anything else. China's Greater Bay Technology says its first all-solid-state A-sample cells have passed nail-penetration and crush tests. The company is aiming for GWh-scale mass production and use in vehicles in 2026.23 Japan's programs follow a slower schedule. Toyota is targeting limited-production EVs in 2027–2028, and it began building a solid-electrolyte pilot plant with Idemitsu Kosan in January.24 Samsung SDI is sending pilot-line samples to automakers and targets mass production in the second half of 2027.24 In China, CATL, BYD and Gotion are working on several electrolyte types at once, with demonstration vehicles generally expected around 2027.26

The sceptical coverage is persuasive. One industry analysis says any EV marketed as "solid-state" in 2026 actually uses a semi-solid hybrid.27 It puts all-solid pilot-line yields below 70% and estimates new production lines at up to $112 million per GWh.27 Another guide says no fully solid-state battery is in any production EV sold in the US. It does not expect affordable mainstream deployment before 2030–2032.28 Even the specifications vary between reports. Toyota's first-generation range appears as about 1,000 km in one account and as a projected 1,200 km for a prototype in another.2124 QuantumScape's 1,000-cycle result is reported as more than 95% energy retention in one source and more than 80% capacity retention in another.2428

On the evidence, semi-solid cells are the solid-state product that is actually shipping now. NIO, for example, already sells a 150 kWh semi-solid pack.25 All-solid cells look set to reach premium flagships first, with volume production unlikely before the end of the decade.

5. Lithium-sulfur: right chemistry, different vehicle

Lithium-sulfur has the highest theoretical ceiling of the five, often quoted above 500 Wh/kg, and it needs no nickel or cobalt.33 Its weakness is still cycle life. Lyten, widely seen as the most advanced startup, has a target of 500 cycles.39 Lyten's chief marketing officer has called lithium-sulfur "a multi-decade battery" that is at "year four or five." He said this year's focus is drones and autonomous systems.36

Lyten's biggest move this year was buying Northvolt's Swedish assets, including 16 GWh of capacity in Skellefteå.38 But the plant will restart by making conventional NMC lithium-ion cells while lithium-sulfur is industrialised.38 That says a lot about how close the chemistry is to cars. For now, lithium-sulfur looks like a battery for aviation and defence, with EVs a later goal.32

What it adds up to

Across all five technologies, the pattern is consistent. The chemistries arriving first are the ones that fit existing factories and target a specific weakness: cold-weather performance and cost for sodium, energy density for silicon, affordable range for LMR. The bigger leaps, all-solid-state and lithium-sulfur, still need new manufacturing that has not yet proven its yields or costs.

Geography matters too. China is leading on sodium and on the volume of solid-state pilot programs.26 Western companies are betting on silicon and LMR partly to build supply chains less dependent on China.442 For buyers, the next real improvements in range and price are more likely to come from these incremental chemistries than from a single breakthrough cell. The rest of this decade looks set to run several chemistries side by side, each matched to a different kind of vehicle.