Every BYD sold outside China runs lithium iron phosphate. Most German and Korean EVs, and the long-range Teslas, run a nickel cathode. Owners are told one is safer and the other better, which is not what the difference is. The two chemistries trade against each other on five axes, and a buyer who knows which axis they care about can tell in advance which car will annoy them.
This is the chemistry comparison. What BYD’s Blade format adds on top of LFP is a separate question, answered in Blade Battery and LFP: what is actually in a BYD pack.
What it is#
Both are lithium-ion. The lithium, the electrolyte and the graphite anode are broadly the same. The cathode differs, and it sets almost everything else.
| LFP | NMC | |
|---|---|---|
| Cathode | Lithium iron phosphate, LiFePO₄ | Lithium nickel manganese cobalt oxide |
| Critical metals | None: iron and phosphate | Nickel, manganese, cobalt |
| Nominal cell voltage | 3.2 V | 3.6–3.7 V |
| Specific energy | 95–172 Wh/kg; newer cells 180–205 | above 300 Wh/kg at cell level |
| Cycle life | 2,500–9,000, conditions dependent | about 1,000–2,300 |
The lower cell voltage is the root of the density gap. Energy is voltage times charge, so an LFP cell of the same size holds meaningfully less energy before anything else is considered.
How it works#
Energy density, the axis LFP loses#
An LFP pack is heavier and bulkier for the same kWh. That is the entire cost of the chemistry, and everything below is what the money buys.
It matters less in a car than the cell numbers suggest, because a pack is not only cells. LFP’s tolerance for being packed tightly, with no module casings, recovers much of the loss at pack level. That is what cell-to-pack designs are for, and it is why an 82.5 kWh LFP pack fits under a mid-size SUV floor at all.
Price, the axis LFP wins hardest#
On those averages an 82.5 kWh pack is worth about $6,700 in LFP and about $10,600 in NMC. No maker passes the whole difference to the customer, but it is why the cheap end of every range went LFP first, and why the same money buys more kWh in a BYD than in a nickel-based rival.
Cycle life, and what a cycle is worth#
A cycle is one full charge and discharge, not one plug-in. Counting cycles alone flatters LFP, because a cycle is not the same distance in both cars.
| LFP example | NMC example | |
|---|---|---|
| Car | Sealion 7, 82.5 kWh usable | Model Y Long Range RWD, 74.0 kWh usable |
| Real consumption | 196 Wh/km | 157 Wh/km |
| Distance per full cycle | about 420 km | about 470 km |
| At the low end of the cycle range | 2,500 cycles, about 1,050,000 km | 1,000 cycles, about 470,000 km |
Even at the pessimistic end, both numbers are far beyond the life of the car. Cycle life is not what limits an EV battery in normal use. Calendar ageing is.
Calendar life, where both chemistries are alike#
A pack degrades on the calendar whether or not it is driven, and the two things that drive it are the same for both chemistries: state of charge and temperature.
Two things follow, and owners get them wrong in both directions:
- The damage is done by sitting at a high charge in the heat, not by passing through it. Charging to 100 % and driving off is not the same event as parking at 100 % for a fortnight in August.
- LFP’s advantage here is real, but it is not immunity. Storage state of charge and temperature couple the same way in LiFePO₄ cells.
Charging, where the rules genuinely differ#
This is the one place where copying advice across chemistries does harm.
An LFP cell’s voltage is nearly flat between roughly 20 % and 90 % charge, so the battery management system cannot read state of charge from voltage in the middle of the range and has to count charge in and out instead. That error accumulates. Only at the top does the voltage rise sharply enough to re-anchor the estimate and balance the cells. A nickel cell’s voltage falls steadily across the whole range and needs no such appointment.
Both makers say so in their own words:
Cold, and safety#
LFP accepts less current when cold and reaches its peak DC rate later, so a winter fast charge is slower. Preconditioning before a DC stop is worth more on LFP than on nickel.
On thermal behaviour, the iron-phosphate bond holds its oxygen more tightly, which is why LFP resists thermal runaway. The dramatic nail-penetration numbers in circulation are manufacturer tests on manufacturer samples. The chemistry advantage is real; the specific figures stay marketing until somebody independent repeats them.
What it means for owners#
If your car is LFP — every BYD sold outside China, the cheaper Teslas, most Chinese-built entry models:
- Charge to 100 % as a matter of course, at least weekly. The gauge needs it more than the cells mind it.
- Do not leave it parked full in the heat for weeks. Daily 100 % is fine. A month at 100 % in a summer car park is not.
- Expect a slower fast charge in winter, and precondition if the car offers it.
- Expect a range readout that jumps after a full charge. That is the estimator correcting itself, not capacity returning.
If your car is nickel-based:
- 50 to 90 % daily, 100 % before a long drive, in the manufacturer’s own words.
- You carry less weight per kWh and use less energy per km for it.
- Your usable daily range is not the brochure number. It is that number times your charge limit, which is the whole subject of Sealion 7 against Model Y.
Warranty is where a maker’s confidence in its own chemistry shows. BYD warrants the Blade pack in Europe for eight years or 250,000 km to a minimum 70 % state of health. Tesla warrants the Model Y pack for eight years or 160,000 km. A mileage limit is the most honest degradation claim a maker makes, because it has to pay for it.
What we have not verified#
This is an explainer built from published sources. Nothing in it is our own measurement.
- No cell data sheet for any BYD or Tesla cell is public. Every density and cycle-life figure here is a chemistry-level range, not a pack in a car.
- The BloombergNEF prices are survey averages across all applications and buyers. What BYD pays itself for its own cells is not visible.
- We have measured degradation on no car, in no climate, over no period.
- The Battery University storage table is general lithium-ion and says so. We have found no equivalent published table measured on automotive LFP.
- Second-generation Blade, announced 5 March 2026, is a different pack — but not, as far as anyone has published, a known chemistry. BYD gave an energy-density gain of 5 % and a charging curve and disclosed no cathode or anode. The LMFP-plus-silicon-carbon description in circulation traces to vendor write-ups with no BYD statement behind it; this page asserted it until 20 September 2026 and no longer does. It ships on Chinese-market models, and we have seen no evidence of it in a European or Moroccan car.