Tesla's Q2 2026 Earnings and AI/Robotics Pivot

Tesla's capex more than doubled in the quarter - $5.79-5.8 billion, up 142% year-over-year - and the company reaffirmed full-year 2026 guidance above $25 billion, roughly triple what it spent in 2025^[1]. That spending is going into Robotaxi, Optimus, AI compute infrastructure, chip fabrication (TERAFAB) and solar manufacturing, and management expects it to keep rising over the next two to three years^[1]. The immediate cost showed up in the cash flow statement: free cash flow turned negative for the first time in more than two years, landing around -$1.1 billion, even as revenue hit a record $28.24 billion^[2].

Musk's own framing of the tradeoff was blunt on the call - he said Tesla should spend on capex as fast as it can without being wasteful, and that it's 'okay to be a little less capital efficient' if it means getting Optimus, Robotaxi and AI compute built sooner^[1]. Wall Street's response was to cut price targets: JPMorgan, Cantor Fitzgerald and Mizuho Securities all trimmed their numbers after the report, and Tesla posted its lowest operating profit in six years despite the record top line^[3]. Truist's William Stein kept a Hold rating, describing the AI progress as 'positive, but imperfect', while Direxion's Ryan Lee said the core unresolved question for investors is simply whether any of this AI spending will monetize on a timeline that matters^[2]. The bull case and the bear case aren't disagreeing about the size of the bet - they're disagreeing about whether Tesla can afford to wait for it to pay off.

Optimus has missed essentially every deadline Tesla has set for it. In 2022, Musk promised the robot would be 'production ready' by 2023 with thousands running in factories by year-end - it wasn't^[4]. In 2025, the target was 5,000 units; Tesla delivered only a few hundred, a shortfall of more than 90%^[4]. By January 2026, Musk described the deployed units as still 'primarily for learning, not productive tasks,' calling the program 'very much in the R&D phase'^[4]. On the Q1 2026 call in April, Musk set a new target - production starting 'late July or August' at Fremont - while also calling it 'literally impossible' to forecast the ramp given roughly 10,000 unique new components in the robot^[5]. Two days before the Q2 call, Optimus's production count still sat at zero^[6].

Against that backdrop, Musk told the Q2 call Optimus 'will be the biggest product ever' and that the manufacturing ramp will follow a normal S-curve - flat and long at first 'because of the newness of the parts and the robot'^[7]. The pattern across four years is consistent: bigger promises, later delivery, and each miss gets reframed as the next quarter's inflection point.

Tesla's headline numbers looked strong - record deliveries, revenue up 26% - but the underlying auto economics weakened. Average selling price fell to $42,730 from $45,345, automotive gross margin excluding credits dropped to 16.3% from 19.2% in Q1, and regulatory credit revenue - a pure margin boost with no delivery cost attached - collapsed to $146 million, down roughly two-thirds year-over-year^[2]. Combined with the capex surge, that's how Tesla ended up posting its highest-ever quarterly revenue alongside its lowest operating profit in six years^[3].

The reported GAAP net income figure itself drew scrutiny beyond the standard earnings coverage: retail-investor discussion pointed out that a large share of that net income reflected an unrealized, non-operating gain from revaluing Tesla's stake in SpaceX rather than profit from selling cars or software. That distinction matters because it means the operating business - the part that actually funds the AI buildout - looks weaker than the topline net income suggests, reinforcing the same tension analysts like Ryan Lee raised about whether Tesla's core businesses can sustain years of $25 billion-plus annual capex without external cash cushions.

Musk used one line on the call to summarize where he thinks Tesla's value now lives: 'They're actually buying Tesla Full Self-Driving with a car attached, as opposed to a car with FSD'^[7]. The numbers back the framing directionally - FSD active subscriptions reached about 1.48 million, up 56% year-over-year, more than 55% of new North American deliveries now include FSD at purchase, and cumulative FSD miles are approaching 12 billion^[8]. Robotaxi, meanwhile, expanded to seven U.S. metro markets including three in Florida, with Musk describing growth as 'literally exponential' alongside what he called an 'impeccable' safety record^[7].

Musk tied this all together by describing Tesla's capex-funded buildout across Optimus, Robotaxi, AI compute and chip fabrication as 'probably the fastest industrial scale-up since World War II in America'^[9]. That's the pitch investors are being asked to buy into: not an automaker with a software add-on, but an AI and robotics company that happens to sell cars today. Truist's Stein effectively endorses that reframing, valuing Tesla more on FSD, Robotaxi and Optimus than on the auto business itself. But it's also the framing skeptics point to when comparing Tesla's valuation multiple against traditional automakers - a gap that only makes sense if the AI and robotics bet actually monetizes, which is exactly the part still unproven this quarter.