Solid-State Battery Energy Density: How to Read the Numbers

A solid state battery energy density figure is only usable once you attach three labels to it: is it gravimetric (Wh/kg) or volumetric (Wh/L); is it material-level, cell-level or pack-level; and did it come from a lab coupon cell or a production-representative cell. A number without those labels cannot be compared with anything. Two 2026 examples show why: ProLogium's announced 185.4 Ah cell is quoted at 381 Wh/kg and 903 Wh/L, while QuantumScape's QSE-5 B sample is quoted at 301 Wh/kg and 844 Wh/L. Both say "solid-state" and both land near 850 to 900 Wh/L, yet one is about 27 percent higher on Wh/kg than the other.

Energy density is delivered energy divided by something. Gravimetric energy density is watt-hours per kilogram; volumetric energy density is watt-hours per litre. The denominator is where the ambiguity lives: at the material level it is one active material, at the cell level it is the whole cell including current collectors, the separator or solid electrolyte, binder, tabs and packaging, and at the pack level it adds module structure, battery-management electronics, cooling and any pressure hardware. Lithium metal has a theoretical specific capacity near 3861 mAh/g, roughly ten times graphite's practical 372 mAh/g, and a volumetric capacity near 2062 mAh/cm3, about three times graphite. Those material-level numbers set the ceiling, not the cell.

This guide is the deep dive on one line item in our broader method for how to verify a solid-state battery claim: the energy-density number. Everything below assumes you already hold a figure from a supplier or an announcement and need to decide what it means.

A spec sheet that quotes only one axis is not describing the same cell as one that quotes the other. Gravimetric energy density rewards removing mass; volumetric energy density rewards removing volume. A design can win decisively on one and only tie on the other.

The two most-quoted solid-state cells of 2026 make the point. ProLogium reported that a 185.4 Ah large-format cell of its Gen 3.5 Lithium Ceramic Battery reached 381 Wh/kg and 903 Wh/L, with TUV Rheinland reviewing the energy-density figures. QuantumScape reported that the B sample of its QSE-5 cell, an anode-free design with a ceramic solid separator and about 5 amp-hours of capacity, delivered a measured 21.6 watt-hours at a C/5 discharge and 25 degrees Celsius, which works out to 844 Wh/L and 301 Wh/kg.

Read together, those are not contradictory. The two cells are close on volumetric energy density. On gravimetric energy density ProLogium's figure is about a quarter higher, consistent with a different cell construction and mass fraction. If a supplier gives you only the volumetric number, you have learned almost nothing about how heavy their pack will be. Ask for both, measured on the same cell.

Every step from an active material to a delivered pack adds mass and volume that store no energy, so the energy-density number falls at each step.

Material to cell. An analysis of energy densities for sulfide solid-state batteries points out that most published work evaluates energy density using only the cathode active-material mass and ignores the solid electrolyte, current collectors and packaging, which is the main reason reported and practical figures diverge. That work models cell-level systems targeting 300, 350 and 400 Wh/kg with a sulfide electrolyte. For lithium cobalt oxide solid-state cells with an approximately 50 micrometre solid-electrolyte separator, a summary of the same analysis gives roughly 410 to 280 Wh/kg and roughly 820 to 560 Wh/L as the electrode's active-material content falls, against roughly 250 Wh/kg and 700 Wh/L for a conventional graphite lithium-ion cell. Separator thickness alone moves the number by a large margin.

Cell to pack. QuantumScape states that a cylindrical-format cell loses roughly 9 percent of its energy density purely from packing cylinders into a rectangular pack, before any structure or electronics. A single-cell-level assessment in ACS Energy Letters adds that the volume change from plating lithium creates significant strain, up to about 17 percent, in larger prismatic cells, and suggests large lithium-metal packs may need a pressure-management system in the same way conventional packs need a thermal-management system. That hardware is pack mass and volume a cell-level number never sees. A "500 Wh/kg" claim means a very different thing at each level, and a supplier who does not say is leaving off the most important label.

Even at a fixed cell-level boundary, a small laboratory cell and a production-representative cell built from the same materials report different energy densities. A techno-economic assessment in Nature Energy puts it directly: solid-state lithium-metal batteries show promise for gravimetric and volumetric energy densities upwards of 500 Wh/kg and 1,000 Wh/L, but most experimental reports at the cell level fall well short of state-of-the-art lithium-ion because of high solid-electrolyte thicknesses and low areal capacities.

Two levers explain most of the gap. A lab cell often uses a thick, robust electrolyte pellet that is easy to handle but heavy; a production cell needs a thin separator that is hard to make defect-free. And a lab cell often runs a low areal capacity, a thin coating that cycles cleanly but dilutes the active fraction; a long-range fast-charging electric-vehicle electrode needs to reach a practical areal capacity of at least about 7 mAh/cm2 while still accepting roughly a 4C charge, according to the same ACS Energy Letters assessment. The Faraday Institution frames the target from the other side: hitting a 1,000 Wh/L automotive figure constrains the lithium anode reservoir to no more than about 17 micrometres.

This is why company laboratory claims and targets belong in a separate column from measured production cells. The Faraday Institution notes a company laboratory claim of roughly 450 Wh/kg cells and a separate aspirational claim of 1,000 Wh/kg with a ten-year life; both are real statements but neither is the same evidence class as an independently reviewed cell. The lithium-metal negative electrode is the lever behind much of the energy-density gain, but this guide does not treat anode architecture as an independently released topic.

When two suppliers send you energy-density numbers, do not rank them. Run each one through the same six questions, then compare the answers.

Sort the result: accept if both figures survive all six questions on the same basis; needs clarification if a label is missing; not comparable if the two figures sit at different boundaries, stages or conditions and the supplier will not restate them. An open cell-level estimator that includes inactive components, such as the SolidPAC battery-on-demand estimator published in Cell Reports Physical Science, is a reasonable way to put two figures on the same basis when the assumptions differ.

Put the two 2026 cells through the procedure. Axis: both suppliers give both axes, ProLogium 381 Wh/kg and 903 Wh/L, QuantumScape 301 Wh/kg and 844 Wh/L. Boundary: both are cell-level, single announced cells, not packs, and neither figure includes module structure or pressure hardware.

Test article: different. ProLogium's is a 185.4 Ah large-format cell; QuantumScape's is an approximately 5 Ah cell, and larger cells generally carry a smaller inactive fraction. Conditions: QuantumScape states C/5 at 25 degrees Celsius and a measured 21.6 Wh; ProLogium attributes its energy-density figures to a TUV Rheinland review and adds that UL Solutions tested the cell under the China standard GB/T 43568-2026, including a vacuum-and-heat test with under 0.05 percent weight loss against a 0.5 percent threshold, but the public detail on discharge conditions is thinner. Stage: ProLogium said mass production had begun; QuantumScape's number is from a B sample. Who measured it: ProLogium's energy density was reviewed by TUV Rheinland; QuantumScape's is a company measurement.

Verdict: the two cells are comparable on volumetric energy density and both are credible cell-level figures, but they are not directly comparable on gravimetric energy density without normalizing for cell size and stage, and neither is a pack figure. That is a useful, honest conclusion, and the kind you should expect most of the time. For how a normalized solid-state cell figure stacks up against the LFP and NMC cells shipping in volume today, see our solid-state versus LFP versus NMC comparison.

If you are scoping solid-state cells for a program, do not ask "what is your energy density." Send your requirement on a fixed basis: the target gravimetric and volumetric figure, the boundary you need it at (cell or pack), the areal capacity, charge rate, temperature window and stack pressure your application allows, and the sample stage you need to see. Then ask every supplier to answer on that basis.

Review the battery product families for format and voltage context, then use the technical contact path to request samples and align on a test protocol. AntBattery can discuss a project-specific cell and supply an evidence package on the same measurement basis. No single cross-chemistry energy-density figure is implied by this article.