TechNewsFirst

In our September 19 piece on Tesla's subscription move, we benchmarked FSD against Adobe, Amazon Prime and the self-storage industry. The structural argument: Tesla had switched from a high-price, low-volume one-time sale to a low-price, always-renewable recurring bill. Once buyers were locked into the payment cadence, the model would compound. The $10 trillion market cap call rested on that compounding, not on FSD being a perfect product.

Q3 2026 data puts a number on how fast that conversion is moving.

According to Ashok Elluswamy, Tesla's VP of AI Software, 75 percent of new Tesla deliveries in North America included an FSD subscription in Q3 2026. Up from 55 percent in Q2. A 20-point jump in a single quarter. Tesla reported total global subscribers at 1.74 million, up 67 percent year-over-year. More than half are on subscriptions rather than one-time purchases, meaning roughly 875,000 people are paying $99 a month. That annualizes to just over $1 billion in recurring FSD revenue.

For context: when we wrote the September piece, Tesla had just disclosed at Q2 earnings that new subscriptions were outpacing one-time purchases for the first time. The Q3 number is what that trend looks like one quarter later.

The speed matters for the thesis. Our Adobe comparison was about a forced transition: Creative Suite buyers howled, signed a petition, but most eventually converted. Tesla's conversion looks less like coercion and more like preference. Elluswamy specifically noted that customers subscribe "even though they have that free month available." Every new Tesla gets a one-month free trial. They are subscribing before they have had to decide. That is not a captive customer tolerating an increase. That is a customer deciding the $99 is worth paying from day one.

The Amazon Prime playbook from the September piece needs revisiting too. We argued that a subscription, once established, gives the seller small, infrequent, below-threshold increases that compound over time. The 55-to-75 percent jump in one quarter suggests Tesla is still in volume-acquisition mode, not margin-extraction mode. The mechanism for incremental price increases is sitting unused while attach rates climb.

Here is the one thing that could make the 75 percent headline mislead. Elluswamy disclosed no churn data. "Subscribed at pickup" and "still paying at month twelve" are different numbers. A 75 percent attach rate that fades to 40 percent after six months tells a much weaker subscription story than the current headline suggests. Tesla has not released retention figures and was not asked about them in the reporting InsideEVs sourced this from. That gap is the load-bearing unknown in the thesis right now.

The 12 percent figure from the September piece, the share of all Tesla owners who had ever paid for FSD in any form, is being repriced rapidly by every delivery that comes in at a 75 percent attach rate. The fleet is not static. If the attach rate holds at anything close to current levels for the next several quarters, that lifetime percentage will look unrecognizable within a year.

Our $10 trillion forecast, dated September 19, 2026, rested on the subscription becoming the dominant revenue mechanism for Tesla's software layer, which then justifies the AI-platform valuation multiples Morgan Stanley assigned. One quarter of 75 percent North American attach does not close that case. But the trajectory is moving in the direction the forecast requires, and it is moving faster than the historic comps implied. The churn data, if Tesla releases it, is what we will score this against.

*Q3 2026 FSD attach rate, subscriber count, and Elluswamy quote reported by InsideEVs, October 21, 2026. The Q2 comparison figures and the $10 trillion prediction are from our September 19, 2026 piece, linked in the dossier.*