Tesla spent the quarter ended June 30 proving it can still move cars at scale, then spent the cash those cars produced on an autonomy and robotics build-out that crushed operating leverage. Deliveries recovered after last year's volume-crossover factory changeover, Full Self-Driving attach rates rose in markets where the software is approved, and energy storage returned to a heavy deployment quarter. The income statement did not celebrate. Operating income fell even as revenue rose, because research spending, stock-based awards tied to the chief executive performance plan, and a capital budget now guided above twenty five billion consumed the volume rebound.
The tension is quality of earnings versus quality of story. Automotive sales grew because cash deliveries grew, not because regulatory credits returned. Credit revenue fell to $146 million from $439 million a year earlier. Services and other finally looks like a profit pool rather than a fleet support cost. Energy deployed 13.5 gigawatt-hours and then gave the margin back through a vendor-cell warranty charge, vanished first-quarter tariff benefits, and lower industrial storage prices. Below the operating line, a mark-to-market gain on a newly purchased SpaceX stake made net income look healthier than the factory floor.
The next year resolves whether Robotaxi miles, Cybercab production, and paid Full Self-Driving subscriptions start to fund themselves, or whether the car business remains a cash engine asked to underwrite science-project economics. Active Full Self-Driving subscriptions reached 1.48 million as Robotaxi service ran unsupervised in six cities. Operating margin compressed to 1.4 percent even after record second-quarter deliveries. Cybercab public rides opened in Austin in early September, and federal safety staff opened a certification inquiry the same week. Does unsupervised ride-hailing become a real profit line before the spending cycle leaves operating income thinner for longer?