Tesla’s second-quarter results, released on July 22, delivered a split verdict for the global EV industry: sales recovered, but profitability weakened sharply as the company entered what Elon Musk called its biggest-ever investment cycle. At the same time, Volkswagen deepened its China strategy by handing a key part of its Level 3 autonomous driving roadmap to local supplier Horizon Robotics, while Porsche approved a broader restructuring plan amid weak EV momentum. Together, these developments show an industry moving into a more demanding phase—one where scale alone is no longer enough, and success will depend on software, localization, and capital discipline.
Tesla: Record Revenue, Shrinking Profit
Tesla posted a strong top-line quarter but a much weaker earnings profile.
Key Q2 figures
- Revenue: $28.24 billion, up 26% year-on-year
- GAAP net profit: $1.11 billion, down 5%
- Operating profit: $398 million, down 57%
- Operating margin: 1.4%
- Deliveries: 480,000 vehicles, up 25% year-on-year
- Free cash flow: -$1.09 billion to -$1.1 billion, Tesla’s first cash-burning quarter since early 2024
- Capital expenditure: $5.79 billion, up 142%
- Expected 2025 capex: more than $25 billion
Tesla also said its trailing 12-month revenue topped $100 billion for the first time.
Yet the core problem is clear: more cars did not translate into stronger profits.
| Metric | Q2 2025 | YoY Change | What it means |
|---|---|---|---|
| Revenue | $28.24B | +26% | Demand recovered at the headline level |
| Deliveries | 480,000 | +25% | Best-ever second quarter for volume |
| Operating profit | $398M | -57% | Profitability deteriorated sharply |
| Operating margin | 1.4% | Down sharply | Core earnings under pressure |
| Auto gross margin | 16.3% | Below expectations | Pricing and mix remain challenging |
| Average revenue per vehicle | $42,730 | Down from $45,345 | Discounts and lower-priced models weighed |
| Regulatory credit revenue | $146M | Down from $439M | Less support from non-core income |
| Free cash flow | -$1.1B | Turned negative | Growth is being funded with cash burn |
The market’s interpretation is straightforward: Tesla used lower pricing, incentives, and entry-level products to restore volume, but at the cost of per-vehicle revenue and margin.
The Biggest Spending Cycle in Tesla’s History
Musk’s message on the earnings call was that Tesla is no longer optimizing for short-term capital efficiency. Instead, it is spending aggressively to build the infrastructure for its next businesses.
The investment areas include:
- AI compute infrastructure
- Robotaxi deployment
- Cybercab production
- Optimus humanoid robot manufacturing
- Chip development and manufacturing
- Factory expansion
- Solar and energy capacity
That helps explain why investors focused less on this quarter’s delivery rebound and more on the timeline for monetization. Tesla still has cash, and the free-cash-flow miss was not as bad as some feared, but the central question is now whether these investments can produce revenue quickly enough to justify the scale of spending.
FSD, Robotaxi and Cybercab: Momentum, But Limited Proof
Tesla continues to frame the car as a software platform rather than just a vehicle.
What Tesla reported
- FSD active subscriptions: 1.48 million, up 56% year-on-year
- In North America, more than 55% of new vehicle deliveries reportedly included FSD subscription at handover
- Robotaxi has expanded to 7 US metropolitan areas
- The fleet has completed about 380,000 miles of unsupervised driving, according to management
- Cybercab has started production at Gigafactory Texas
- Tesla says it has installed manufacturing capacity for 125,000 Cybercabs per year
But the details investors want are still missing:
- Actual Robotaxi fleet size
- Daily ride volume
- Revenue per vehicle
- Remote assistance rate
- How many FSD users are paying full price versus using free trials or bundled offers
- Real Cybercab output versus designed line capacity
This is an important distinction. Coverage across more cities demonstrates geographic adaptability, but it does not yet prove commercial scale.
Tesla also acknowledged a notable strategic pause: Musk said it would be unreasonable to deploy unsupervised FSD or Robotaxi at large scale when a major software architecture upgrade is imminent and could improve safety. That is a more cautious tone than in prior quarters and suggests Tesla is prioritizing safety validation and regulatory resilience over aggressive rollout headlines.
Optimus: Expectations Are Being Reset
Tesla’s humanoid robot program remains central to Musk’s long-term narrative, but management is clearly dialing back near-term expectations.
According to Tesla:
- The first-generation Optimus production line is being installed
- Production is expected to begin within this year
- Early robots will be used internally for data collection and function development
- Initial units will not be aimed at external customers
Musk described Optimus as possibly the hardest product Tesla has ever tried to scale, noting that there is no mature off-the-shelf supply chain for humanoid robots. He also warned of “a few difficult quarters ahead.”
That matters because embodied AI and robotics have been surrounded by fast-rising expectations in 2025. When the industry’s most visible benchmark company openly slows the timetable and emphasizes manufacturing complexity, it is signaling that robotics may follow a much longer S-curve than many investors had hoped.
Volkswagen’s China Bet: Horizon Robotics Gets the L3 Mandate
While Tesla is spending heavily to build its own stack, Volkswagen is moving deeper into a localized China strategy.
On July 22, Volkswagen China’s smart driving unit Carizon expanded its cooperation with Horizon Robotics. The two companies aim to co-develop a unified end-to-end driving model based on Horizon’s AI foundation, running from raw sensor input directly to driving decisions.
The target
- Level 3 autonomous driving in China
- Planned for H2 2027
Before that
Carizon’s existing advanced driver assistance system will begin rolling out from Q3 this year across:
- Volkswagen’s three joint ventures in China
- 7 new EV models
This is strategically significant. A major global automaker is effectively entrusting the algorithmic foundation of its China intelligent-driving roadmap to a domestic supplier. It reinforces a broader trend in the Chinese EV market: foreign brands increasingly need China-based software, China-trained models, and China-specific execution to remain competitive.
| Company | Partner/Approach | Target Timeline | Strategic Meaning |
|---|---|---|---|
| Tesla | In-house FSD/Robotaxi/Cybercab stack | Ongoing, cautious expansion | Full-stack vertical integration but high capex and execution risk |
| Volkswagen China | Carizon + Horizon Robotics | L3 in H2 2027 | Localization through Chinese ADAS/AI expertise |
Porsche’s Restructuring Shows the Pressure on Legacy Premium Brands
Another signal from Europe came from Porsche, where the supervisory board approved a new round of restructuring. According to the source, Porsche plans to expand job cuts to 9,000 positions, roughly doubling previous reduction plans.
The backdrop is a sharp decline in profits and weak momentum in parts of its EV lineup.
Reported restructuring points
- Previous plan already targeted 3,900 job cuts over several years
- Management now reportedly wants to add 5,000 more, taking the total to 9,000
- Separate reports suggest 5,000 to 6,000 more cuts by 2035 may also be considered
The report also ties Porsche’s actions to Volkswagen Group’s wider cost-cutting program, which has included:
- Potentially cutting up to half of existing model lines by 2030
- Reducing option complexity by 75%
- Lowering annual capacity from 10 million vehicles to 9 million
Most strikingly for EV watchers, the report says the Porsche Taycan may exit after its current lifecycle without a direct successor.
If that proves accurate, it would underline a harsh reality for some traditional premium brands: electrification alone does not guarantee a viable product strategy, especially when Chinese EV makers are moving faster on software, value, and update cycles.
Why This Matters
These three stories point to a common industry shift.
1. EV competition is no longer just about volume
Tesla delivered nearly half a million vehicles in the quarter, but margin compression dominated the narrative. The market is rewarding credible software and autonomy monetization more than raw sales growth.
2. China is becoming the center of intelligent-driving execution
Volkswagen’s Horizon deal shows that in the Chinese EV market, local autonomous-driving partners are increasingly essential. This is not just supplier localization; it is strategic dependence on China’s AI and mobility ecosystem.
3. Legacy European brands are under dual pressure
Porsche’s restructuring highlights the squeeze from both sides:
- Slower-than-expected EV returns in some premium segments
- Intensifying pressure from Chinese EV makers on price, tech, and product cadence
4. The capital cycle is getting tougher
Tesla’s more than $25 billion capex plan and negative free cash flow show how expensive the next stage of EV and autonomous competition has become. The winners may be those with enough balance-sheet strength to endure a multi-year payoff period.
Global Implications for the EV Market
For global automakers, the message is increasingly clear:
- Autonomy is becoming region-specific, especially in China, where local regulations, data, and road scenarios favor domestic technology ecosystems.
- Software revenue is central to EV valuations, but investors are demanding harder evidence on paid adoption, retention, and unit economics.
- Manufacturing scale is no longer enough if it is not matched by battery security, chip access, and efficient capital deployment.
- Chinese suppliers are moving up the value chain, from components to core AI and autonomous-driving architecture.
This matters beyond China. If Volkswagen’s China-first localization model works, other global brands may follow with deeper partnerships in ADAS, cockpit AI, and vehicle operating systems. At the same time, if Tesla proves that FSD, Robotaxi, and Cybercab can scale into a profitable software-led mobility business, it could reset investor expectations across the entire EV sector.
What to Watch Next
Over the next 12 to 24 months, several metrics will matter more than headline deliveries:
For Tesla
- FSD paid conversion and renewal rates
- Robotaxi fleet size, utilization, and safety disclosures
- Cybercab actual production output and operating economics
- Optimus internal deployment scale and supply-chain progress
- Whether free cash flow remains negative as capex ramps
For Volkswagen in China
- Speed of ADAS rollout across the 7 new EV models
- Whether Carizon and Horizon can deliver a credible path to L3 by 2027
- How effectively Volkswagen can localize software while protecting brand differentiation
For Porsche and legacy European brands
- Whether cost cuts translate into stronger EV profitability
- Product-line simplification and model rationalization
- How they respond to Chinese EV competition in both China and export markets
The broader takeaway is that the EV industry is entering a less forgiving phase. The easy growth story has given way to a capital-intensive battle over software, autonomy, and supply chains. Tesla is betting that massive spending today will create a defensible mobility and robotics platform tomorrow. Volkswagen is betting that in China, local intelligence is the only viable path forward. Porsche’s restructuring, meanwhile, is a reminder that even premium brands are not insulated from the new economics of electrification.
In short, the next winners in the EV market may not be the companies that sell the most cars this quarter, but the ones that can turn technology ambition into durable, profitable execution.
