Referral value as a percentage of the deal, Model 3 Standard to Model Y L

"Is a referral code worth it" is a ratio question that almost never gets answered as one. The numerator is fixed; the denominator is whatever car you are buying. Here is the arithmetic against 2026 US prices.

The numerator

In North America the buyer benefit is 3 months of Full Self-Driving (Supervised). FSD has been subscription-only here since 14 February 2026 at $99 a month, so the trial displaces about $297 of spending. For a prior Enhanced Autopilot owner paying the $49 rate, the same three months displaces about $147 — the derivation is in the $49 rate calculation.

The table

Top to bottom, the same benefit is worth 2.7 times as much proportionally on the cheapest referral-eligible car as on the most expensive one.

What the ratio is good for

Two decisions, and not the one people reach for. It is not useful for deciding whether to use a code: the answer there is always yes, because it costs nothing. It is useful for sizing the benefit against the other things on the order sheet. At $36,990, $297 is comparable to a paint option and worth a moment's thought. At $99,990 it is inside the noise of a delivery fee.

And it is useful for calibrating how much effort to spend chasing one. Reorganising a purchase, delaying an order, or accepting a worse configuration to secure a referral benefit is irrational at every price in the table. Spending forty seconds opening a link before you configure is rational at all of them.

Why the cheapest car makes the benefit most visible

There is a behavioural point behind the arithmetic. At $36,990 a buyer is typically choosing between the base configuration and one upgrade, so a $297 benefit lands in the same mental bucket as a paint option — a real, comparable line. At $99,990 the same $297 disappears into the rounding of fees and taxes, and no buyer at that price is making a decision differently because of it. Same benefit, same programme; the only thing that changed is what the buyer is comparing it against.

Two things the percentage hides

First, the benefit is a subscription trial, not a discount: it does not reduce the price, the financed amount, or the tax base — so "0.80% of the deal" describes value received, not money saved on the invoice. Second, it converts to a paid $99 monthly subscription at the end of month three unless you act, which is a cost the percentage does not show at all. The cancellation path is covered in where a trial gets cancelled.

Outside North America the ratio is not calculable this way

Mainland Europe's referral benefit is Supercharging distance — roughly 2,000 km since April 2026 — whose cash value depends on local per-kWh pricing and on whether the holder can even reach the stalls before the six-month clock runs out. A percentage would be false precision, so this site does not publish one.

Related: the full method in the referral savings formula.