Tesla
AI-driven electric vehicles, Full Self-Driving (FSD), and commercial solutions for humanoid robots
The world's largest real-world driving data loop + end-to-end AI model (FSD V12) + vertically integrated manufacturing capability, building dual moats in autonomous driving and robotics
Company Overview
Tesla was founded in 2003 (Musk took over in 2004) and went public on NASDAQ in 2010, with its headquarters located in Austin, Texas, USA. The company's core businesses encompass electric vehicles, energy storage systems, and AI-driven autonomous driving and humanoid robot solutions. Within the AI industry chain, Tesla is positioned at the industry application layer (B2B), adopting an end-to-end large model (FSD V12) as its core technical roadmap, expanding AI from perception to a closed loop of decision-making and control, while progressively extending into the commercialization exploration of general-purpose robots (Optimus).
Tesla
Tesla is the world's largest pure electric vehicle manufacturer, delivering approximately 1.79 million vehicles globally in 2024. Core models include the Model 3/Y (volume models, accounting for ~95% of total deliveries) as well as the high-end Model S/X and Cybertruck. Leveraging its self-developed three-electric system (battery, motor, and electronic control), giga-casting, and 4680 battery technologies, the company continues to reduce manufacturing costs. The EV business serves as a massive data collection platform for AI systems, capturing real-world road data at scale.
FSD (Full Self-Driving) is built on an end-to-end neural network architecture. In 2024, Tesla officially rolled out V12 to North American users, eliminating traditional rule-based code and enabling direct camera-input vehicle control. As of early 2025, FSD has accumulated over 3 billion miles driven, with a subscription price of approximately $99/month (US) and a one-time purchase option of $8,000. This business is transitioning from one-time sales toward a high-margin subscription model, with commercial deployment of Robotaxi expected to begin in 2025.
- Energy Storage: Powerwall/Megapack, with over 20 GWh deployed in 2024, benefiting from growing electricity demand from AI data centers.
- Optimus Humanoid Robot: Began performing simple material handling tasks inside Tesla factories in 2024, with a target price below $20,000. Long-term scale could reach tens of billions of units, but the product remains in early R&D/testing stages and has not yet contributed meaningful revenue.
| Product Line | Revenue Share | Core Customers | Gross Margin |
|---|---|---|---|
| Electric Vehicles | ~85% | Global consumers, government/corporate fleets | ~18.5% |
| FSD/Autonomous Driving | ~8% | Tesla owners, Robotaxi operators | ~70%+ (extremely low marginal cost) |
| Energy Storage | ~6% | Residential, commercial & industrial users, utilities | ~15% |
| Robotics (Optimus) | <1% | Internal factories (pilot phase) | Negative (R&D investment phase) |
Technical Moat
Moat 1: The World's Largest Real-World Driving Data Closed Loop
Tesla has a fleet of over 7 million vehicles equipped with FSD hardware, generating billions of miles of real-world road test data daily. Through distributed training on the Dojo supercomputer (self-developed D1 chip), the model iteration speed far exceeds that of competitors. The data is not only used for autonomous driving but also provides a foundation for generalized scenario learning for Optimus.
Moat 2: End-to-End AI Model (Pixels in, Actions out)
FSD V12 completely abandoned the intermediate rule layer, adopting a single massive neural network to directly complete the end-to-end mapping of perception → planning → control. This architecture eliminates scenario shortcomings caused by hand-crafted rules, and the model can be trained in parallel across thousands of GPUs, forming a "data flywheel." Competitors (such as Waymo) still rely on HD maps and rule engines, making it difficult to replicate Tesla's scale of iteration speed.
Moat 3: Vertical Integration and Manufacturing Capability
Tesla has self-developed FSD/Dojo chips, 4680 batteries, and electric drive systems, and has built ultra-large integrated die-casting production lines. This vertical integration deeply couples AI systems with hardware, reducing dependence on external supply chains, while providing a low-cost sensor, motor, and battery reuse foundation for new businesses such as Optimus.
| Dimension | Data |
|---|---|
| Global EV market share | Approximately 18% (2024) |
| Autonomous driving road-test mileage | Over 3 billion miles (cumulative FSD), ranked first globally |
| Industry ranking | No. 1 in EV sales; top tier in autonomous driving technology |
| Main competitors | Waymo (Alphabet), Cruise (GM), BYD, XPeng Motors, UBTech (robotics) |
| Downstream customers | End consumers, corporate fleets, ride-hailing platforms, industrial manufacturing enterprises |
Financials and Growth
| Metric | Data |
|---|---|
| Revenue (2024) | USD 96.77 billion |
| Gross Margin | Approximately 18.2% (automotive business) |
| Net Margin | Approximately 10.3% |
| Core Growth Logic | 1) FSD subscriptions and Robotaxi contribute high-margin service revenue; 2) Optimus robot opens up a trillion-level market; 3) Energy storage business benefits from surging electricity demand driven by AI computing power |
Key Risks:
- Regulatory Risk: Autonomous driving regulations may progress slower than expected, especially in the EU and China, potentially delaying the large-scale commercialization of Robotaxi.
- Data Security and Privacy: Large-scale collection of road video data faces cross-border data transfer restrictions in various countries; China data must be processed locally.
- Intensified Competition: Waymo, Cruise, BYD, Huawei, and others are accelerating their efforts in autonomous driving and robotics, and technology roadmaps may diverge.
- Slowing Vehicle Demand: The growth rate of global EV penetration is slowing; Tesla needs to rely on AI service revenue to offset declining hardware margins.
Core Investment Logic / Industry Value Summary:
Tesla is the most aggressive practitioner in the application layer of the AI industry. It has disrupted the traditional autonomous driving paradigm with end-to-end models and replicated this approach to humanoid robots. Its real-world data closed loop plus vertical manufacturing capabilities form a moat that is difficult to replicate. If FSD/Robotaxi and Optimus succeed commercially, Tesla will transform from an automobile manufacturer into an AI service and robotics platform company, unlocking a multi-trillion-dollar incremental market.