LiFePO4 Battery Price Index 2026

LiFePO4 prices fell again in 2026, but the shape of the price curve has changed. Cell prices are close to the floor set by lithium carbonate and processing cost, so the savings now come from pack architecture — larger cells, fewer parts, higher-voltage stacks — rather than from cheaper cathode material. The result is a market in which two quotes can differ by 40% per kWh and both can be honest, because they describe different objects: a bare cell, a wall-mounted pack with a five-year warranty, or an outdoor cabinet with HVAC and aerosol fire suppression.

The practical consequence for buyers is that comparing US dollars per kWh without a bill of scope is meaningless. A 100 kWh outdoor cabinet at $150/kWh and a 100 kWh rack of modules at $95/kWh are not the same product. One includes thermal management, smoke detection and suppression, an IP55 enclosure, a two-tier battery management architecture and a factory acceptance test. The other is a stack of cells in a metal box.

This index exists to make that difference legible. It publishes one band per product class, on one consistent basis, and shows the arithmetic that connects them — so that a procurement team, a journalist or an analyst can check a quotation against a reference instead of against another quotation.

The index is compiled from two independent inputs. The first is DLXN’s own 2026 price book: 37 active SKUs across cells, home packs, floor-standing systems and commercial cabinets, quoted ex-works in China across the year. For each SKU we take the low and the high quote observed during 2026, convert them to US dollars per kWh of nameplate energy, and trim the top and bottom decile of the resulting distribution, so that one distressed order or one sample-unit quotation cannot move the published band.

The second input is the publicly reported spot market for prismatic LFP cells in the 280–320 Ah format. Cell spot quotes are the most liquid reference point in the chain. They set a hard floor under pack pricing, because no pack can sell below the cost of the cells inside it plus assembly, and they are quoted daily rather than quarterly. We use the spot range to sanity-check the bottom of every band; we do not use it to derive the band, because system pricing includes work that a cell quote never sees.

Every figure is factory-gate, ex-works China, per kWh of nameplate energy measured at 25 °C and 0.5C discharge. Currency conversion uses DLXN’s 2026 planning rate for CNY/USD, applied identically to both inputs, so the published ranges move with the exchange rate but the internal relationship between them does not. Bands assume orders of at least ten packs or one cabinet; single-unit and sample orders typically quote 10–20% above the top of the band.

We publish ranges rather than point prices for four reasons. Cell grade, A versus A-minus, moves cell cost by roughly 8–12%. BMS architecture — passive versus active balancing, and whether the unit speaks CAN to a specific inverter brand — moves pack cost by 10–15%. Enclosure material and ingress rating move cabinet cost by 6–10%. Warranty length and certification scope move every category. A single number would average those differences into a fiction, and a fiction cannot be cited.

The index is revised quarterly. This edition uses data through 1 November 2026 and will be superseded in January 2027. Historical bands stay on the same URL, so a citation made today still resolves to a table with the revision recorded, rather than to a dead page.

The table below is the core of the index. Each row names a product class, the configuration the band refers to, and what the band includes. Read the last column before comparing rows: the step from $75 to $110 per kWh between a cell and a commercial cabinet is not a margin story, it is a scope story.

Cell pricing sets the floor. At $55–75/kWh, one 314 Ah cell at 3.2 V (about 1.0 kWh nominal) costs $55–75 before grading, matching and delivery. Packs are priced per kWh of installed energy, so the same cell contributes $55–75 of material cost while the finished pack sells at $95–135. The gap between those two numbers is the subject of the next section.

The stack below is the average composition of a 100 kWh commercial cabinet at the midpoint of its band, US$110/kWh. We publish it because the most common procurement error in storage is comparing a cell quote with a system quote as if they described the same thing. The cell is 62% of the system; the other 38% is what turns cells into an asset that can be installed, insured and warranted.

Purchase price per kWh is a capital metric. What a battery actually delivers is energy throughput over its service life, and that is the number to compare against a grid tariff, a diesel generator or a demand charge.

Take a 10 kWh home pack at $95/kWh, the bottom of the residential band, with 80% usable depth of discharge and 8,000 cycles. The pack costs $950. Over its life it moves 10 × 0.8 × 8,000 = 64,000 kWh, so each kWh of throughput costs $950 ÷ 64,000 = $0.0148. That is about 1.5 US cents. Applying a 90% round-trip efficiency raises it to $0.0165.

At $110/kWh the same pack costs $1,100 and delivers the same 64,000 kWh, so throughput costs $0.0172/kWh, or $0.0191 after efficiency. At $150/kWh, the top of the commercial band, the figures are $0.0234 and $0.0260. The entire spread from the cheapest to the most expensive LiFePO4 on this page is therefore 1.5 to 2.6 cents per kWh cycled — much narrower than the 3× spread in headline capital cost suggests.

Two cautions. First, 8,000 cycles at one cycle per day is 21.9 years, longer than most warranties on this class of product. The honest planning number for a warranted asset is the cycle count in the warranty: DLXN home packs are warranted at 6,000 cycles at 80% DoD, and at $110/kWh that raises throughput cost to $0.0229/kWh, a third higher. Second, this formula covers the battery only. Where the battery is 60% of a hybrid system’s cost, a whole-system figure is roughly 1.6× the numbers below, before inverter efficiency.

The commercially meaningful comparison is with the alternative source. In off-grid and weak-grid markets that alternative is usually diesel, whose cost per kWh is dominated by fuel and logistics rather than capital; in grid-connected markets it is the retail tariff and the demand charge. Storage wins where the daily price spread or the avoided fuel cost exceeds roughly 2–3 cents per kWh plus the cost of capital — and in 2026 a growing number of markets clear that bar.

Every number so far is ex-works China. Landed cost is a different page of the invoice, and the difference is large enough that a European buyer comparing a factory quote with a distributor quote may conclude that one of them is wrong. Neither is: the spread is duty, certification, freight and working capital.

We model four destination regions as a premium over the ex-works band. These premiums are not different list prices — they are the cost of putting the same cabinet on a site in that region, including the compliance work that a local installer will not do for free.

Cell format. The 280–320 Ah prismatic cell is the 2026 workhorse, and 314 Ah and larger formats are displacing 280 Ah in new designs because fewer cells per pack means fewer busbars, fewer sense lines and less assembly labour. Expect the cell band to drift toward its floor as the larger format scales, without any change in cathode chemistry.

Cell capacity utilisation. LFP cell capacity added in 2024 and 2025 still exceeds demand, which is why cell prices have stayed near the carbonate-plus-processing floor. A demand recovery in utility-scale storage would tighten this first, and pack prices would follow within one to two quarters.

Certification and insurance. Fire-test evidence is becoming the binding constraint in more markets than cell supply is. A cabinet with cell-level UL 9540A data can be specified into projects that a cheaper cabinet cannot enter at any price — which is why the cheapest quote is frequently not the lowest landed cost.

Tariffs. Trade policy is now a larger source of price variance than manufacturing cost in North America. Any 2027 forecast that assumes a stable duty environment should be read as a scenario, not a base case.