When I wrote my first piece on Donut Lab’s solid-state battery, the main reason I thought the story had become worth following was simple: it had started moving beyond launch-day claims and into public, third-party-backed testing.

The newest development is no longer just another cell-level headline. Donut Lab has now also been covered for an actual pack-level charging test inside a Verge motorcycle, and that changes the discussion more than the third self-discharge result by itself.

What the third test was about

The third test focused on self-discharge, which is a much less glamorous topic than charge speed or high-temperature performance. According to Donut Lab’s public summary, VTT connected the cell to a battery tester for ten days and measured voltage every ten seconds. The company says the voltage stabilized within the first ten hours and then stayed level for the remaining nine and a half days. (I Donut Believe)

That matters because the test is aimed directly at a basic skepticism that naturally shows up around bold battery claims: if something charges unusually fast and behaves differently from conventional lithium-ion cells, is it really acting like a battery, or is it behaving more like a supercapacitor?

Donut Lab’s summary frames the result as evidence of normal battery-type charge retention, not the rapid linear discharge you would expect from a supercapacitor. (I Donut Believe)

Why this matters more than it first appears

This is not the kind of result that gets turned into a flashy CES headline.

But it is exactly the kind of result that matters when you are trying to decide whether a battery story is moving from “interesting marketing” toward “worth taking seriously.”

Fast charging gets attention. High-temperature behavior gets attention. Self-discharge sounds boring. But if a product failed this kind of test, it would immediately undercut the whole conversation about real-world usability.

A battery that cannot retain charge predictably is not just inconvenient. It changes whether the thing behaves like a practical energy-storage system at all. So while the third test does not prove the biggest claims, it does help close off one of the simpler ways the story could have unraveled.

What the three public tests now cover

Taken together, Donut Lab’s public VTT-backed tests now cover three different parts of the credibility picture:

  • Fast charging: the company says VTT confirmed 0 to 80% charge in 4.5 minutes at 11C without active cooling. (I Donut Believe)
  • High-temperature behavior: the company says the cell delivered about 110% of nominal room-temperature capacity at 80 C and about 107% at 100 C while remaining rechargeable afterward. (I Donut Believe)
  • Self-discharge: the company says the voltage remained stable over a ten-day idle test, supporting the case that the cell behaves like a battery rather than a supercapacitor. (I Donut Believe)

That is still not the same thing as proving the entire Donut Lab story. But it is more substance than most breakthrough battery stories ever show publicly.

The 12-minute charging detail now looks more serious

This is the part that materially changes the story for me.

Earlier, the strongest charging evidence was still cell-level. Donut Lab’s first VTT-backed charging announcement said the cell reached 100% state of charge in just over 12 minutes at 5C, with the full charged capacity available on discharge afterward. That already mattered because it suggested the battery was not just accepting a quick surface charge and then fading immediately on discharge. (Donut Lab)

But the more important update is the newer pack-level charging test inside a Verge TS Pro motorcycle. Electrek reported that the standard 18 kWh pack sustained more than 100 kW of charging power for five straight minutes, moved from 10% to 50% in five minutes, reached 70% in just over nine minutes, and hit 80% in 12 minutes, all in an air-cooled configuration. Interesting Engineering covered the same development as another step beyond the earlier cell-only demonstrations. (Electrek, Interesting Engineering)

That does not settle the full Donut Lab debate, but it is a more meaningful result than a lab-bench cell test for one simple reason: it shows the conversation moving into pack-level behavior in an actual vehicle.

I still think the wording matters, though. There are three different layers of evidence here:

  • the cell-level case, which is getting stronger through VTT-backed measurements
  • the pack-level case, which looks stronger now that a Verge motorcycle pack has been shown charging at real power levels
  • the commercial case, which is getting bolder through Verge’s product claims and launch messaging

Those layers are connected, but they are not interchangeable. A pack charging result in a real motorcycle is a more serious milestone than a cell-only test, but it is still not the same thing as comprehensive independent validation of the whole Donut Lab claim stack, especially around energy density, cycle life, cost, and manufacturing scale.

What still is not proven

I do not think this latest result changes the biggest unresolved questions.

The most ambitious claims are still the ones that need the most proof:

  • 400 Wh/kg
  • 100,000 cycles
  • scalable manufacturing
  • cost claims versus conventional lithium-ion

Those are still the hard parts.

The Donut Lab battery page continues to present the broader vision in very aggressive terms: extremely high energy density, five-minute charging, long cycle life, lower cost than lithium-ion, and broad product applicability beyond motorcycles. (Donut Lab) The new self-discharge result helps with credibility, but it does not independently settle the larger performance and commercialization story.

My take now

My opinion is a little stronger than it was in the first post, but not dramatically different.

I still do not think the full Donut Lab case is proven.

I do think it is becoming harder to dismiss the company as just another battery press-release machine.

Three public VTT-backed tests are not the whole answer, but they are enough to show a pattern: the company is at least trying to expose important parts of the story to external measurement instead of relying only on launch-stage promises. The “I Donut Believe” site also makes it clear that the testing program is still ongoing, which is exactly what I would want to see if the company is serious about narrowing the gap between claims and evidence. (I Donut Believe)

The separate Verge commercialization push makes the story more important, but it also makes it more important to keep the categories straight. Verified cell tests, pack-level charging results, product launch claims, and long-term production validation are not the same thing, and I think the most useful way to watch Donut Lab is to keep insisting on that distinction.

So my view is still basically this:

Donut Lab is worth watching not because the future is settled, but because the story keeps surviving the kinds of checks that would be easy to fail if it were pure hype.