The CATL-made lithium iron phosphate (LFP) battery fitted to the entry-level grade of Tesla Model 3 maintained an average state of health of 93.3% among a fleet of vehicles that had driven over 100,000 km. This is according to a compilation of actual vehicle inspections published on July 8, 2026, by Carla, a Swedish used-EV sales company, and it exceeds the 88.2% average for the Panasonic-made 52.4kWh pack by 5.1 percentage points. The gap observed within the same vehicle body overlaps with laboratory findings suggesting that LFP tends to have a longer service life. That said, the publicly available data falls short of what would be needed to conclusively attribute the difference to battery chemistry alone.
Comparing Four Groups Beyond 100,000 km, a Gap of Up to 5.1 Points in Average SoH
Carla analyzed 9,954 battery inspections conducted in Sweden between 2022 and 2026. Among these, for vehicle models and battery specifications with over 100,000 km driven and a sufficient number of inspections, current capacity relative to initial capacity was compared as State of Health (SoH). The results, broken down by battery for the Model 3, are as follows.
| Battery Specification | Cathode Chemistry | Average SoH Beyond 100,000 km |
|---|---|---|
| CATL 60.5kWh | LFP | 93.3% |
| LG Chem | NMC | 91.5% |
| Panasonic 77.8kWh | NCA | 89.8% |
| Panasonic 52.4kWh | NCA | 88.2% |
The gap between the top and bottom groups is 5.1 percentage points. LFP also outperformed LG Chem's nickel-manganese-cobalt (NMC) chemistry by 1.8 points. Because this comparison is confined within the Model 3, which shares a common body design and basic thermal management, it captures the influence of battery specification more clearly than comparing across different manufacturers' models.
On the other hand, retained "rate" and remaining energy amount are two different things. Even if the 60.5kWh LFP pack maintains a higher percentage, that does not mean its driving range exceeds that of the 77.8kWh Long Range pack. When choosing a used vehicle, one must look not only at the degradation rate but also at initial capacity, vehicle efficiency, and the required single-charge driving range.
Carla's overall ranking also includes a figure of 92.83% for the 78.8kWh LG Chem pack, narrowed down to a vehicle group with "newer software." The population appears to differ from the 91.5% figure used in the per-battery comparison, but the relationship between the two is not explained. When reusing such figures, one cannot mix vehicle-group definitions simply because the battery manufacturer name is the same.
What LFP Gains in Longevity, and What It Loses in Return
In the olivine-type crystal structure of the LFP cathode, strong P–O covalent bonds support the framework. Materials research published in Inorganic Chemistry links this bond to high thermal safety and reports that LiFePO4 exhibits excellent reversibility in lithium insertion and extraction. While this is a material property consistent with durability, it is not evidence that chemistry alone explains the inter-group differences observed by Carla. According to Tesla's own materials, LFP packs use neither nickel nor cobalt.
The trade-off is energy density. In its 2021 Impact Report, Tesla outlined a strategy of allocating NCA and NMC to high-energy applications and LFP to low-energy applications. Master Plan Part 3 similarly states that standard-range vehicles require LFP, while long-range vehicles require high-nickel cathodes. The fact that the LFP-equipped Model 3 showed higher SoH does not negate the value of the long-range specification. For use cases prioritizing distance and weight, the higher energy density of nickel-based chemistries remains advantageous.
Also incorrect is the understanding that LFP cells are unaffected by being left at full charge. A 2025 study examining LFP/graphite cells found that as storage temperature rose from 25°C to 50°C and state of charge increased from 50% to 100%, capacity loss accelerated, with temperature having the larger effect. Tesla itself no longer specifies a single uniform charge limit, instead instructing owners to follow the battery-specific recommended values displayed on the vehicle's screen.
Even with nickel-based chemistry, degradation varies depending on usage. A study of Panasonic-made NCA cells extracted from a 2018 Model 3 found that cycling repeatedly within a narrow state-of-charge range at very high or very low levels accelerated degradation, and that during storage, higher states of charge led to greater capacity loss. Chemistry is one factor affecting battery life, but temperature, state of charge, and charging power also govern the rate of degradation.
Even with 9,954 Inspections Across All Models, a Chemistry-Based Victory Cannot Be Confirmed
The AVILOO FLASH Test used by Carla acquires up to roughly 1,000 signals from a stationary vehicle in about three minutes. It reads cell voltage and temperature, and also statistically evaluates the battery management system and charging history. This is not a test that merely reads the manufacturer's dashboard display. However, it is also not a test that directly measures actual capacity by fully charging and then discharging the battery.
AVILOO states, based on reference measurements against the PREMIUM Test—which uses a complete discharge cycle—that at least 95% of cases fall within ±3 percentage points of error. This represents the accuracy of testing a single vehicle, not the uncertainty inherent in a 5.1-point average difference between vehicle groups. Because Carla has not disclosed the number of vehicles and average vehicle age for each battery group, statistical significance cannot be determined. The distribution of mileage beyond 100,000 km, as well as variance and confidence intervals, also remain unknown.
Vehicle age matters in particular. Even with the same mileage, a car that traveled long distances over a short period differs from one that reached the same mileage over many years in terms of the ratio between charge-discharge-driven degradation and calendar-time-driven degradation. Battery manufacturers are also tied to production year and capacity, and manufacturing location and software are not necessarily identical either. The four groups examined here do not constitute a controlled experiment in which all conditions besides chemistry were held equal. Nor can average values obtained under Sweden's climate be directly transferred to vehicles in hotter regions.
Caution is also needed regarding the fact that this is a compilation of vehicles selected by a sales company. Carla explains that it inspects vehicles for sale that are compatible with AVILOO, and if a vehicle fails, it is repaired or has its battery replaced before being re-inspected. From the published materials, it is unclear how the initial failing data or replaced packs were incorporated into the aggregate figures. Therefore, this result shows that "among inspected used Model 3 vehicles, the average SoH of the LFP group was higher," but it does not measure failure rates or the lifespan of all vehicles.
From Mileage to SoH: The Next Standard for Used EV Appraisal
For buyers of used EVs, the practical value of this data lies not in the label "LFP" but in the necessity of individual inspection. On Tesla vehicles, one can check whether a car is equipped with an LFP battery via "Controls," "Software," and "Additional Vehicle Information." For compatible vehicles, connecting to an AC charger of 5kW or more from below 20% state of charge and running the Battery Health Test—which can take up to 24 hours—measures energy retention relative to when the vehicle was new.
Warranty terms should also be checked. The current warranty for the Swedish market covers the Model 3 RWD for 8 years or 160,000 km, and the Long Range and Performance for 8 years or 192,000 km, with a minimum capacity retention of 70% guaranteed within that period. The average for the over-100,000 km vehicle groups compiled by Carla significantly exceeds this standard, but an average does not guarantee any individual vehicle. Only by aligning SoH, battery replacement history, and remaining warranty period on a per-vehicle basis can price differences be properly evaluated.
Operational history also factors into price. In a telematics analysis of over 22,700 vehicles updated by Geotab in January 2026, the average degradation rate was 2.3% per year. Vehicles with heavy reliance on DC fast charging above 100kW degraded at up to 3.0% annually, while those relying mainly on AC or lower-power charging degraded at about 1.5%, with hot climates adding a further 0.4 percentage points per year. Even with the same battery type, charging patterns and climate widen the gap.
In the European Union, moves are underway to institutionally close this information gap. The EU Battery Regulation has required, since August 2024, that battery management systems retain up-to-date data for determining the state of health of EV batteries. For EV batteries placed on the market or put into service on or after February 18, 2027, a battery passport will be required, recording chemistry, individual state of health, and charge-discharge cycle counts, among other data. This is not a system that automatically applies retroactively to existing used vehicles, but the era of pricing based primarily on mileage and model year is drawing to a close.
The LFP group's average of 93.3% can serve as material for checking battery specification at the time of purchase. To support a final judgment, Carla would need to next disclose the number of vehicles and vehicle age for each battery group, show the distribution of mileage and how replaced vehicles were handled, and confirm whether the 5.1-point gap holds under the same model year and usage conditions.