How much has your battery actually degraded? Feed in charge sessions (battery % before → after, kWh added) and get the same median-pack estimate, care score, and 3-day streak alerts the app computes — no install, no car, no account.
DEMO DATA — pretend car, pretend sessions. This page never connects to a real Tesla.Method: a session from A% to B% adding K kWh implies a pack of K ÷ ((B − A) ÷ 100) kWh. Sessions with swings under 20 points are ignored — dividing by a tiny swing magnifies rounding error. The estimate is the median of the rest, so one odd session can't move it. Degradation = 1 − (estimated ÷ nominal).
| Date | Location | Start % | End % | kWh | Sat at end (h) | Implied pack |
|---|
Try it: change the Sep 10–12 rows so the car no longer sits at 100% for 8 hours — watch the streak alert disappear and the care score climb.
The full app (github.com/rtmalikian/tesla-battery-guru)
records real sessions from your Tesla (or ones you type in), keeps them on your own
computer, and exposes an OpenAI-compatible coach API
(/v1/chat/completions, /v1/models, /v1/embeddings)
so any AI client can answer "how degraded is my battery?" from your own numbers.
It never commands a car and sends your data only to the AI backend you configure —
or nowhere at all, with the built-in rules engine.