Google Maps, Apple Maps, PlugShare, and A Better Routeplanner all assume your EV charges at its advertised peak speed, drives at its EPA-rated efficiency, and fills up linearly like a gas tank. For a lot of real vehicles — conservative Battery Management Systems, no active preconditioning, an aging second-hand pack — none of that is true. ChargeCliff exists to show the physics those tools quietly assume away.
26 vehicles and counting — from the conservative-BMS Lexus RZ and Toyota bZ4X to Tesla's better-behaved curve to the Nissan Leaf's famously bad one, plus trucks, luxury sedans, and everything between. If your car's real-world charging curve has ever made you feel lied to by its spec sheet, this is built for exactly that moment.
It's also a portfolio piece: a small, fully first-principles physics engine with every formula derived. Seven profiles are calibrated against a named, published dyno or road-test benchmark; the rest are considered estimates from public charging reports, clearly labelled as such in the API and the UI — not a black box, and not a guess dressed up as more than it is.
Every charge curve in this tool was tuned by running the actual Riemann-sum session simulation against a real published number — a factory dyno spec, a winter road test, a magazine's 75mph range test — and adjusting breakpoints until the model landed inside that published range. Those checks are pinned as automated regression tests, not a one-time claim: a future change that moves a result outside its published range fails the test suite.
A profile with no named benchmark (the Toyota bZ4X, Subaru
Solterra, and VW ID.4 today) is marked verified: false
in the API and flagged in the UI — a platform-sharing estimate,
presented with exactly the confidence it's earned, not more.
A DC fast-charge session integrated across discrete 1% State-of-Charge intervals — not a single kWh-divided-by-kW shortcut, because the charge curve genuinely isn't flat. Includes a cold-gating derate below 68°F and reports the exact SoC where sitting at the charger longer stops being worth it.
Aerodynamic drag scales with the square of speed; real highway cruising (70-75mph) plus a resistance-heater cabin warmer in cold weather can cut EPA-rated range 35-50%. Reports both the theoretical full-charge range and the standard 80%→10% SoC road-trip window everyone actually plans around.
A State-of-Health estimate from age and odometer alone — no OBD-II dongle, no manufacturer app — and the compounding power derate an aged pack's BMS applies on top of its own factory curve.
Per-minute billing charges the same rate regardless of how fast power is actually flowing, which means a slow-tapering EV can end up paying more per kWh — sometimes more per mile than gasoline — on the exact same billing plan as a fast-charging 800V car.