Tesla’s AI bet needs robotaxis to work
- Tesla still earns most of its revenue from cars, but the stock story is now autonomy.
- Cybercab robotaxi rides are live for paying customers in Austin, with Dallas and Houston next.
- Nearly 1.3 million customers have paid for FSD, giving Tesla a real software base.
- Energy storage is profitable, but large Megapack sales to Musk linked entities add concentration risk.
- The One Big Bill and tariffs are pressuring EV demand, credits, battery costs, and margins.
- Tesla plans over $25B of capex across 2025 and 2026, so free cash flow may stay under pressure.
Autonomy upside, near term pressure
The bull case is that Tesla is becoming an AI transport company, not only a car maker. It has launched paid robotaxi service in Austin with Cybercab, and the service is expanding to Dallas and Houston. Full Self Driving, or FSD, has nearly 1.3 million paid customers globally. If FSD becomes truly unsupervised, Tesla could earn software and ride revenue from a fleet it already helped place on the road.
Energy storage is the other strong piece. Megapack and Powerwall help grids and homes store electricity, and the segment is still a profitable growth engine. Tesla also has a large AI plan. Management expects more than $25 billion of capital spending across 2025 and 2026 for six factories, compute, and chip work, including the Terafab effort with SpaceX and Intel.
The bear case is serious. Tesla’s 2025 revenue fell by $2.86 billion to $94.83 billion, and net income attributable to common stockholders fell by $3.30 billion to $3.79 billion. The One Big Bill repealed the $7,500 U.S. EV credit and removed emission standard penalties, which hurts demand and regulatory credit revenue. Tariffs are also raising costs in both automotive and energy.
The key question is timing. Optimus production is expected to start around late July or August, the new Roadster may debut soon, and wider FSD v14.3 or v15 releases could show whether the autonomy story is getting closer. But the valuation gives Tesla little room for slow progress. The company needs robotaxis, AI infrastructure, energy growth, and robots to turn into real cash flows, not only exciting demos.
Cars fund the AI buildout
Today Tesla makes money mostly by building and selling electric vehicles. The Model 3, Model Y, and Cybertruck drive the current auto base. In Q1 2026, the automotive segment reported $19.98 billion of revenue, compared with $2.41 billion for energy generation and storage. Automotive margins excluding credits improved to 19.2 percent in Q1 2026, which shows better cost control even as policy pressure grows.
The long term plan is to move from one time vehicle sales to Transport as a Service. That means Tesla wants cars to drive people around for pay, like an Airbnb on wheels. FSD subscriptions and purchases are the bridge. Robotaxi service in Austin is the first public proof point, but regulators and safety records will decide how fast it can spread.
Energy storage gives Tesla a second growth engine. Megapack serves utilities and large customers, while Powerwall serves homes. The 2025 10-K/A disclosed that xAI bought $430.1 million of Megapack products in 2025 and $78.1 million through February 2026. The Q1 2026 10-Q also showed $87 million of revenue from SpaceX Megapack purchases. These are real sales, but they also tie part of energy growth to related parties in the Musk ecosystem.
Tesla is spending heavily to protect its AI future. It recorded $390 million of automotive segment expenses in the second half of 2025 for supercomputer assets, contract terminations, and AI restructuring. It also invested $2.00 billion in SpaceX common stock in Q1 2026. The bet could create a wide lead in autonomy, chips, and robots. The risk is that shareholders fund years of spending before the returns arrive.
From EVs to robots
Model 3 and Model Y
These are Tesla’s core mass market vehicles. They still carry most of the car volume and help fund autonomy, AI compute, and new factories.
Cybertruck and Cybercab
Cybertruck expands Tesla into pickups. Cybercab is the dedicated robotaxi platform, with initial production started and paid rides live in Austin.
Model YL, Model Y Performance, and Model 3/Y Standard
These newer trims were launched to fill more factory capacity and reach more buyers. They can support volume, but lower price points may weigh on average selling prices.
Model S and Model X
These older premium models are being wound down. Tesla plans to convert their Fremont factory space into an Optimus factory with a long term goal of one million units a year.
Semi
Semi targets freight customers that want electric heavy trucks. Management expects early online builds next year and real volume in the back half.
Roadster
The new manually driven Roadster is expected to debut soon. It is more of a brand and technology showcase than a near term profit driver.
Megapack, MegaBlock, and Powerwall
Megapack and MegaBlock serve grid scale storage, while Powerwall serves homes. This business is profitable, but battery costs, tariffs, and related-party concentration matter.
Optimus
Optimus is Tesla’s autonomous humanoid robot. Production is expected to start around late July or August, but the product is still early and demand is unproven.
Two reportable segments
Tesla’s Q1 2026 Form 10-Q reports two segments: automotive and energy generation and storage. Based on Q1 2026 segment revenue, automotive was about 89 percent of reportable segment revenue and energy was about 11 percent, with energy sales partly supported by related-party Megapack purchases.
What could break the thesis
Robotaxis fail to scale
High impact · Medium oddsTesla’s long term value depends on unsupervised FSD and robotaxi service spreading far beyond Austin. The technology still has to prove it can work safely in many cities, weather conditions, and edge cases. Regulators may also slow or block launches outside early Texas markets.
Policy and tariff shock
High impact · High oddsThe One Big Bill removed the $7,500 U.S. EV credit and cut emission standard penalties to zero. That weakens buyer incentives and threatens regulatory credit revenue. Tesla also says tariffs and OBBBA provisions could raise battery cell costs and hurt demand for energy products.
AI spending outruns cash returns
High impact · Medium oddsTesla expects more than $25 billion of capex across 2025 and 2026. The money is going into factories, AI compute, and chip work. The 2025 10-K warns that AI needs far more compute, memory, energy, and cooling, which may be too hard or too expensive to secure.
EV price pressure gets worse
Medium impact · High oddsEV competition is intense, and rivals keep discounting to fill factories. Tesla has launched lower priced variants to increase utilization and expand its market. That can help deliveries, but it can also push down average selling prices and margins.
Related-party reliance grows
Medium impact · Medium oddsTesla disclosed $430.1 million of 2025 Megapack revenue from xAI and $78.1 million through February 2026. The Q1 2026 filing also showed $87 million of revenue from SpaceX Megapack purchases, and Tesla invested $2.00 billion in SpaceX common stock. These links may help Tesla build a shared AI and energy ecosystem, but they also raise questions about customer concentration and capital allocation.
Product goals miss real demand
Medium impact · Medium oddsTesla’s 2025 10-K warns that goals tied to the 2025 CEO Performance Award may not match future customer demand. That matters because Tesla is shifting factory space and capital toward Optimus, robotaxis, and AI. If customers prefer cheaper cars, better service, or different energy products, Tesla could miss more profitable paths.
In one breath
Where does Tesla make most of its money today?
Tesla still makes most of its revenue from the automotive segment. In Q1 2026, automotive segment revenue was $19.98 billion, while energy generation and storage was $2.41 billion.
Why does FSD matter so much for Tesla stock?
FSD is the bridge from selling cars to earning software and ride revenue. Tesla has nearly 1.3 million paid FSD customers and a paid robotaxi service in Austin, but the technology and rules still need to prove they can scale safely.
What did The One Big Bill change for Tesla?
It repealed the $7,500 U.S. EV credit and removed emission standard penalties. That can hurt buyer demand and reduce regulatory credit revenue, while related rules and tariffs may also raise battery costs.
Is Tesla’s energy business clean growth?
Energy storage is profitable and growing in importance, but it is lumpy and exposed to battery costs. It also includes large related-party Megapack sales to Musk linked entities such as xAI and SpaceX.