Tesla’s Q2 2026 Delivery Release, for the Record

Tesla's investor-relations release of 2 July 2026 recorded 480,126 deliveries for the second quarter, up roughly 25% year on year, with European volume up about 40%. This entry records the figure and states what it does not contain.

What a delivery release does not contain

Any referral data. No count of referred orders, no credit issuance, no cap utilisation. Every public claim about referral volume is inference, and this desk treats it as such rather than dressing it up. It is also why "the referral programme is being wound down" and "the referral programme is booming" both appear each quarter with equal confidence and equal absence of evidence.

Different units of account

Company deliveries are counted per quarter. The referral cap is counted per referrer, per calendar year, at the first 10 successful referrals. Those units do not meet: no amount of company volume raises an individual ceiling, and a referrer who reached ten in June gains nothing from a record December.

The 2026 pairing worth recording

Volume rose while the programme narrowed. Model S and Model X were removed in March 2026, and the eligible list for a UK buyer is now Model 3 and Model Y. Anyone reasoning that strong sales imply a more generous incentive has the sign backwards for this year.

The European note, read from the UK

European volume up roughly 40% in the quarter landed in markets where the referral reward is Supercharging distance and where that reward doubled to about 2,000 km in April 2026. The UK figure — approximately 650 miles, about 1,046 km — did not move. Regional aggregates therefore describe a group of markets with materially different rewards, which is why a UK reader should discount them.

Why inference from volume is unreliable in both directions

Two arguments circulate whenever a delivery figure is published. The first holds that strong sales let a manufacturer reduce incentives, so a referral benefit should be expected to shrink. The second holds that strong sales indicate confidence and capacity, so benefits should be expected to expand. Both are plausible, neither is evidenced, and 2026 provides a counter-example to the simpler version of each: the programme narrowed in March while volume grew through the first half. A reference desk records the figure and declines the inference.

The one variable a buyer controls

Whether a referral is attached at all. It attaches when the order is placed, through the link, and cannot be added afterwards. In any market condition that is the single saving which does not depend on timing anything.

What would count as evidence

For anyone attempting to reason about referral activity from public data, the honest answer is that no adequate source exists: Tesla publishes neither referral counts nor credit issuance, and third-party estimates are built from surveys and inference rather than disclosure. A reference entry therefore records the delivery figure, notes what it omits, and stops. Readers wanting a defensible number about their own position have exactly one source, which is their own account.

Verification

Delivery figures are published by Tesla's investor-relations page; programme terms are on the Tesla UK Refer and Earn page. Neither is a substitute for the order summary, which is the only live statement of what a given configuration will receive.