Donut Lab’s Solid-State Battery Passes Its First Independent Test — But the Hardest Claims Are Still Unproven

The VTT Technical Research Centre of Finland has independently verified that a Donut Lab solid-state cell charged from 0 to 80% in 4.5 minutes at an 11C rate — a result that legitimately puts the Finnish startup ahead of every other solid-state developer on raw charging speed. Having tracked solid-state battery announcements for years, I’ve learned to read the fine print. This result is real. It’s also carefully scoped.

• The Fact: VTT report VTT-CR-00092-26 confirms a 26 Ah Donut Lab cell reached 80% charge in 4.5 minutes at 286A (11C), retaining 98.4–99.6% of stored energy on subsequent discharge.
• The Catch: During a single-heat-sink 11C test, the cell surface hit 90°C and forced a test abort — contradicting Donut Lab’s “no active cooling needed” marketing. CES 2026 showed no working prototype, only 3D-printed mockups.
• Still Unverified: The claims that drew the harshest industry criticism — 400 Wh/kg energy density, 100,000-cycle lifespan, and cost parity with lithium-ion — have zero independent confirmation.
• The Deadline: CEO Marko Lehtimäki promised Verge Motorcycles deliveries in Q1 2026. Early orders are reportedly still tracking for late March; new orders have slipped to Q4 2026.

What VTT Actually Tested — and What It Didn’t

VTT’s February 23 report (VTT-CR-00092-26) covered one thing: charging performance. The lab tested a single 26 Ah cell using a PEC ACT0550 battery tester in a climate-controlled chamber, running two protocols — 5C (130A) and 11C (286A) — with passive aluminum cooling plates as the only thermal management. Before and after each charge, the cell discharged at 1C to verify full usable capacity.

The 5C results: 80% state of charge in 9.5 minutes, full charge in just over 12 minutes, 100% capacity available on discharge. Solid numbers. The 11C results: 80% in 4.5 minutes, full charge in just over 7 minutes, 98.4–99.6% capacity retention. Also solid.

Then the thermal incident. During the 11C test configured with only a single bottom cooling plate — the setup closest to a real-world minimal thermal management scenario — the cell surface temperature hit 90°C, triggering VTT’s safety cutoff. The test was paused for four minutes of cooling, the cell was strapped more tightly to the heat sink, and then resumed. Donut Lab has consistently marketed this battery as requiring no active cooling. That claim held at 5C and with dual plates at 11C. With single-sided cooling at 11C, it didn’t.

That’s not a deal-breaker. But it is a gap between the marketing and the measured reality that buyers and engineers should note.

One more context point worth flagging: Donut Lab showed no working prototype at CES 2026. The booth displayed 3D-printed mockups. For a company claiming gigawatt-hour-scale production, the absence of a live demo in January was a legitimate red flag — and one that the VTT charging result only partially addresses.

The Claims That Still Need Proof

Donut Lab made six headline claims at CES 2026: 400 Wh/kg energy density, 100,000-cycle lifespan, extreme temperature tolerance from -30°C to above 100°C, lithium-ion cost parity, non-geopolitically sensitive materials, and sub-5-minute charging. VTT has confirmed one of them — the easiest one to demonstrate on a single cell in a lab setting.

The 400 Wh/kg energy density claim is the one that drew the sharpest reaction. Yang Hongxin, chairman and CEO of Chinese battery manufacturer Svolt Energy, told Chinese media that the battery simply doesn’t exist: “That battery doesn’t even exist in the world; all the parameters are contradictory. Any person with even a basic understanding of the technology would think it’s a scam.” That’s a strong statement from someone with real manufacturing credentials — though it’s worth noting Svolt is simultaneously developing its own second-generation solid-state cells targeting the same 400 Wh/kg figure. His conflict of interest cuts both ways.

The cycle life claim is equally extraordinary. 100,000 cycles would be roughly 10 times what automotive lithium-ion batteries deliver. Nobody has asked an independent lab to charge this cell 100 times and measure degradation, let alone 100,000. Fast charging a cell once — or even a handful of times — tells you almost nothing about longevity. Electrek put it plainly: plenty of cell designs can charge fast once. Doing so repeatedly without significant degradation is the actual engineering problem.

Donut Lab’s Transparent Rollout Strategy Is Itself a Signal

Here’s what I find genuinely interesting about this situation: Donut Lab could have submitted everything for testing at once and waited for a comprehensive report. Instead, CEO Lehtimäki handed individual cells to VTT and let the lab publish whatever it found, in sequence, starting with the easiest-to-verify spec. The company has even launched a dedicated site, idonutbelieve.com, to publish VTT results as they arrive.

That’s not how companies behave when they’re hiding something. You don’t invite Europe’s most respected state-owned research organization to run worst-case tests on your cells and publish everything unless you’re confident in what they’ll find. Lehtimäki said as much to Electrive: “If we had presented full third-party validation right away, the controversy wouldn’t have ended — it would have simply shifted. If certain people don’t like the conclusion, they simply move the goalposts.”

It’s a calculated transparency play. Five claims remain unverified and will arrive on Donut Lab’s schedule. That’s both encouraging and, for those watching the delivery deadline, a source of real tension.

EVXL’s Take

I’ve written about solid-state battery timelines repeatedly over the past year — from our December analysis concluding the technology won’t hit mainstream EVs until 2035 to Toyota’s decade of missed targets and Nissan’s survival-mode battery bet. My working rule has been simple: take any solid-state timeline and add five years. Donut Lab is making me reconsider that rule — at least for the charging claim.

The fast charging data is real. An 11C charge to 80% in 4.5 minutes, verified by VTT with a credible test protocol, is a meaningful result. For context, Stellantis and Factorial’s validated cells — which we covered in April 2025 — charged from 15% to 90% in 18 minutes. That was considered impressive at the time. Donut Lab’s 4.5 minutes to 80% is in a different category entirely.

But charging speed was never the claim battery scientists called impossible. Energy density and cycle life were. Neither has been independently tested yet.

My prediction: within 60 days, VTT publishes an energy density report. If that figure comes in at or near 400 Wh/kg, the entire solid-state development timeline for every major automaker — Toyota, Nissan, Honda — gets thrown into question. If it comes in well below 300 Wh/kg, the fraud accusations get louder and Yang Hongxin looks prescient. Either way, watch idonutbelieve.com. The energy density report will tell us far more than this first one did.

One practical note for anyone tracking the Verge Motorcycles delivery situation: Verge CEO Tuomo Lehtimäki confirmed to InsideEVs that orders placed in 2025 are still targeting late March delivery. New orders placed today go to Q4 2026. Around 350 bikes are planned for the full year. Until media test rides happen and customers post real-world data, the battery’s on-road performance — not just its lab charging curve — remains unverified.

Editorial Note: AI tools were used to assist with research and archive retrieval for this article. All reporting, analysis, and editorial perspectives are by Haye Kesteloo.