Skip to content

Independent — not affiliated with or endorsed by Tesla, Inc. Tesla is a trademark of Tesla, Inc.

Tesla Referral Link – 3 months of FSD free (opens Tesla in a new tab)
US Referral offer · United States: 3 months of FSD free · we may earn Tesla credits john580308 Open at Tesla (new tab)

Note · method

When two archived copies of Tesla's referral page disagree, which one is right?

By the TeslaPromoCodes.com desk · published October 17, 2026

Short answer

The newest capture that shows a normal, uncapped page for the right country wins, and we print its date next to the figure. Most disagreements have one of four causes: the offer really changed between the two dates, one page belongs to an owner who hit the yearly cap, the two pages are from different countries, or the visible line and the page data say different things. Each has a tell you can check.

Referral offer · United States

3 months of FSD free

Tesla's page, captured October 4, 2026.

john580308 Open at Tesla (new tab)

Referral link: Tesla applies the benefit when you order through it; we may earn Tesla credits.

Why this happens at all

Tesla has no public price list for referral benefits. The figure lives only on referral pages, one per owner, and each archived copy is a snapshot of one owner's page in one country on one day. Two snapshots can be weeks apart, from different owners, in different language versions of a country's page. So when you pull up a capture that contradicts the matrix, or a capture that contradicts another website, the first job is not to decide who is lying. It is to work out which of these four situations you are in.

Reason 1: the dates differ, and the offer changed

This is the honest kind of disagreement: both captures are right for their day. Our clearest 2026 example is Thailand. Tesla's Thai referral page captured on July 12, 2026 showed ฿8,500 off Model 3 and a Model Y card with nothing under it; the page data held 0 baht for Model Y. The capture of September 29, 2026 showed ฿8,500 off (card reads "Model Y Excl. Launch Series"). Both are genuine Tesla output.

What we cannot tell you is the day it changed. The Archive has no Thai referral captures between July 13 and September 28, 2026, so the change sits somewhere in that window. When that happens we date the new figure by the capture that first shows it and say the window out loud, rather than inventing a precise switch-over date. The same applies to the larger cars in the Gulf: the English-language UAE page captured on June 13, 2026 still listed AED 2,000 off Model S and Model X, and the captures from June 16 did not. That is "listed on June 13, gone by June 16", not "removed on June 15".

Reason 2: a capped owner's page shows nothing

Tesla's terms let each owner refer up to 10 orders per calendar year. Once an owner reaches that, Tesla still serves their referral page but changes the heading to "[name]'s Referral Limit Has Been Reached" and drops the benefit lines. Anyone who opens that capture sees a Tesla page offering nothing, and could fairly conclude the program has ended in that country.

The Netherlands shows how misleading that can be. Dutch captures from August 21, 2026 read "Bestel met de referral van [name]" (order with [name]'s referral) and printed 1,000 free Supercharging kilometres on Model 3 and Model Y. A Dutch capture from September 13, 2026 read "De referral-limiet van [name] is bereikt" (the referral limit has been reached) and printed no benefit lines at all. Tesla did not change the Dutch offer in those three weeks; one owner ran out of referrals. Norway has three capped captures in September 2026 alone. We found capped pages in English, Dutch, Norwegian, Czech, Polish, German, Arabic and Chinese captures, and we skip every one of them when filling the matrix. That is why the Dutch row still shows 1,000 km of free Supercharging, dated August 21, 2026.

The tell: read the page heading before the model cards. If it mentions a limit, that capture says nothing about the country's offer. Pick another capture from a nearby date.

Reason 3: the wrong country, or a different language of the same country

Each country has its own locale code in the address: en_gb is the UK, en_ie is Ireland, en_ae and ar_ae are the UAE in two languages. Belgium, Switzerland, Luxembourg, Canada, Hong Kong, Macau, Puerto Rico and Thailand each have more than one language path. Comparing a UK capture (650 miles) with an Irish one (1,000 km) produces a "conflict" that is only a country difference; see Ireland vs the UK.

Language paths of the same country are a different case. In every country where we read both languages, the benefit matched: French and Dutch Belgium both printed 1,000 km, and the English and Arabic UAE pages both printed 1,000 km. Formatting can differ, though. Italy's January 2026 capture wrote "1.000 Chilometri" and its September capture "1000 Chilometri"; Switzerland mixes "1'000" and "1’000". A different thousands separator is not a different offer, so always compare the number Tesla saves in its page data, not the typography.

Reason 4: the visible line and the page data say different things

Tesla's referral page carries two layers: the headline you see under each model, and a block of product data underneath (per model: a cash value with its currency, a Supercharging distance, or an FSD trial in months). Usually they match. When they don't, the reason is almost always one of these:

  • Capped pages keep their data. On the September Dutch capture above, the data still says 1,000 km per model while nothing is visible. A tool that reads only the data would call that page a live 1,000 km offer. Tesla would not give that buyer anything.
  • "Up to" headings. Australia's page heading reads "[name] Gave You Up To $350 Off a New Tesla", but its Model Y data holds zero and the Model Y card has no benefit line. The heading describes the best case on the page, not every model.
  • Labels that arrive before the money. Malaysia's page data on July 10, 2026 already labelled its Model Y as excluding the Launch Series while still holding RM 0 for it; the RM 1,000 appeared in captures from August 21. A label is not a benefit until a figure sits beside it.

Our rule is that a figure enters the matrix only when the visible line and the data agree. If they don't, the cell says "not verified" until a capture where they do.

A second worked example: Korea, 2025 against 2026

Korea is the extreme case of Reason 1. Captures from May and June 2025 showed ₩330,000 off Model 3, Model S and Model X; every 2026 capture shows nothing on any model. Here the explanation is written down by Tesla: the Korean terms say referral links stopped being active on July 1, 2025. Someone holding a 2025 capture and someone holding a 2026 capture are both reading Tesla correctly, and only one describes today. The full record is in our Korea close-up.

Our decision rule, in order

How we choose which capture fills a matrix cell.
StepQuestionIf no
1Is it the right country code (and does the country's program run)?Set it aside; it belongs to another row
2Does the heading show a normal page, not a referral limit?Skip it; capped pages say nothing about the country
3Do the visible line and the page data agree?Cell stays "not verified" until a capture where they do
4Is it the newest capture that passes 1 to 3?Use the newer one; record the old figure in the changelog
5Is that capture less than 92 days old?Print it, marked "older capture"

Older figures are not thrown away. When a newer capture replaces one, the change goes into the changelog with both dates, which is how the Model S and Model X history in each country survives even though those cards have left the pages.

How you can repeat the check

The mechanics of opening a capture are in how to check an offer yourself, and the full method is on the methodology page. For a disagreement specifically, open both captures side by side and go through the four reasons in order: compare the dates, read both headings, compare the locale codes, and if you are comfortable with raw HTML, compare each visible line with the product data. In our experience one of the four explains the difference every time. If none does, tell us; that is exactly the kind of case we want to re-read.

One last caution: a capture of someone's referral page is evidence, not a way to order. It carries a stranger's code, possibly one that has since hit its cap. Use it to settle the question, then open a link for your own country before you configure. And if a website's figure matches no capture at any date, it is not a disagreement between captures; it is a stale or unsourced claim.

Sources

Related

All notes