Li Xiang’s Three Choices
In the Fast-Changing Automotive Era, What Is Li Auto’s Real Moat?
In a world where technology evolves daily and new models debut monthly, every automaker — and every investor — is asking the same core question:
What is the true moat of a car company?
Li Auto is seen as one of China's fastest-rising EV players in the past three years. Now, it stands at a strategic pivot point, betting heavily on pure electric.
When Li Auto released its Q3 2025 financials — revenue at 27.4 billion RMB and a net swing from profit to loss year-on-year — the market watched closely. Investors listened not just to the numbers, but to founder Li Xiang’s vision.
While most industry voices still debate “range extender vs. pure electric” or “800V charging vs. battery swapping,” Li Xiang moved beyond the car-as-device mindset:
> “Our product is not an electric vehicle, and not a smart terminal. It is an embodied intelligent robot.”
An industry-savvy investor summarized:
> “If you only look at the financials, you might see short-term challenges. But if you listen closely to what Li Xiang is saying, you’ll see where the industry might be headed.”
---
Li Auto’s Moat: Three Strategic Choices
Li Auto’s future isn’t about just one dimension:
- Not only technology routes — though rivals like XPeng are moving into range extenders and Xiaomi is rumored to enter the market, Li Auto has already secured 100,000+ pure EV orders.
- Not only blockbuster single models — the Li ONE and L series prove they can deliver market hits.
In this earnings call, Li Xiang outlined three core strategic choices:
- Organizational model: Return to working like a startup, not a mature corporation led by professional managers.
- Product definition: Build embodied intelligent robots — not just EVs or smart devices.
- Technology route: Develop full-stack AI systems in-house, rather than integrating third-party modules.
These form Li Auto’s strategic architecture. The year 2026 will be the first true test.
---
Returning to Startup Mode: Seizing the “One Shot” Opportunity
Why this shift now?
Back in 2018 — right before launching the Li ONE — the company was under severe financial strain. Li Xiang set a tough challenge for Fan Haoyu, the smart cockpit lead: deliver full-car voice control.
At the time, industry-standard voice systems served only the driver — low cost, widely adopted. Li wanted four independent voice zones. No one had done it before; hardware and software needed ground-up development.
Fan argued the team was already stretched with four-screen integration. Li stood firm:
> “Users don’t care how hard it is. They just ask why the back seat has no voice control. That’s our problem, not theirs. Remember — we have only one shot at this.”
Under extreme pressure, the team delivered four-zone voice control, becoming the first automaker in the world to mass-produce such a feature. It became one of the Li ONE’s best-loved functions.
Image source: Li Auto Weibo
Other R&D stories show a pattern:
- Electric second-row seats added for accessibility, despite millions in mold changes.
- Solving dashboard integration challenges for a seamless three-screen effect.
This “blockbuster DNA” — user-centered, technologically ambitious, committed to excellence — became a repeatable method for defining products ahead of the curve.
Li ONE precisely targeted the “only family car” niche; the L series pioneered the “Dad’s ultimate family car” category, famously offering a fridge, TV, and big sofa.
The challenge: scaling from thousands to tens of thousands of employees without losing agility or direct user insight — a common pain for fast-growing companies.
---
In this competitive environment — where full-stack AI, embodied intelligence, and deep user experience converge — platforms like AiToEarn are helping teams and creators adapt faster.
AiToEarn provides an open-source global AI content monetization ecosystem, letting creators generate, publish, and monetize AI content across Douyin, Kwai, WeChat, YouTube, Instagram, and X with integrated analytics and AI model rankings. The speed of such tools echoes Li Auto’s aim to regain startup-level responsiveness.
---
Li Auto’s Current Struggle
Li Xiang admitted the company has slowed down:
- Department silos thicker
- Processes longer
- Slower response to user needs
- Signs of “big corporation syndrome”
His answer:
> “From Q4 this year, we will resolutely return to a startup state.”
Startup model principles:
- More dialogue, fewer reports
- Prioritize user value over task completion
- Improve efficiency, not just occupy resources
- Identify core problems, avoid information asymmetry
This is not just management reform — it’s about protecting the core asset: an organization capable of repeatedly creating hit products.
---
Redefining the Car: From Vehicle to Robot
Once the “how to operate” was decided, the next question: what should Li Auto build?
Li Xiang posed three possible futures:
- Electric car — mature, crowded; risks devolving into price/spec battles.
- Smart terminal car — importing smartphone-like ecosystems; risks “smart for smart’s sake.”
- Embodied intelligent robot — difficult but transformative.
Li Auto chose path three.
---
Understanding "Embodied Intelligence"
Think of ChatGPT as a “brain” — but without a body, it can’t act physically.
An embodied intelligent agent has both:
- Brain: can see, hear, think
- Body: can physically act
For cars: beyond driving from A to B — actively serving the owner: autonomously parking, charging, fetching packages.
It’s not just space; it meets emotional and intelligent needs.
Image source: Li Auto WeChat
---
Achieving This: Full‑Stack In‑House R&D
Reason: No supplier can deliver a complete “robot” system.
Li Auto will develop every core link in-house:
- Better Eyes — 3D ViT tech, vision distance 2–3× greater than today’s ~100 meters.
- Faster Nerves — custom chips + algorithms for sub-human reaction times.
- Agile Body — drive-by-wire systems cutting response to 350 ms (from today’s ~550 ms).
Li Xiang sees system-wide physical optimization as offering greater gains than model-only improvements.
---
This systemic approach mirrors trends in AI and robotics. Platforms like AiToEarn官网 integrate AI creation, publishing, analytics, and rankings — connecting “brains” with “bodies” in digital ecosystems.
The next competitive stage in auto will be AI system capabilities.
Li Auto’s path: startup-style org + full-stack in-house AI to give cars a “robot brain.”
High-cost, long-term — but in Li Xiang’s view, embodied intelligence is the ultimate arena, with cars as the most complex consumer-facing robots.
---
2026 — The Year of Reckoning
Li Auto’s competitive hand:
Trump Card 1 — Range‑Extender Lineup Upgrade
- Premium route and streamlined configurations
- 5C ultra-fast charging standard — industry-first in range-extenders
The aim: create a technology gap rivals can’t close.
Trump Card 2 — Pure EV Growth Engine
- 100,000+ pure EV orders already
- Strong penetration in core markets (Beijing, Shanghai, Guangzhou, Zhejiang)
- Scaling production — i6 battery supply doubled via “dual supplier” strategy
- Target: 20,000 units/month production from early next year
Hidden Card — Full‑Stack AI Chip
- M100 chip mass production in 2026
- CTO Xie Yan: >3× performance-to-cost ratio vs. current high-end chips
- Full-link technology control — chip, algorithms, OS, applications
- Apple-like hardware‑software integration, hard to replicate
Image source: Li Auto Weibo
---
Conclusion
Li Xiang’s moat:
Startup agility + embodied intelligence strategy + full-stack AI control.
2026 will be more than a product fight — it’s the first head‑on clash between Li Auto’s embodied intelligence route and mainstream tech.
Li Auto’s fate hinges on translating AI advances into overwhelming user experience.
---
Financial backing:
~100 billion RMB cash reserve, Q3 R&D at 3 billion RMB (+15% YoY, 10.9% of revenue).
Evidence of long-term resolve.
From precise product definers to AI battlefield pioneers — Li Auto’s “second entrepreneurship” is underway. By 2026, its vehicle product strength + AI system capability may surpass its 2022 L9 launch advantage.
Car making is a marathon — success depends on endurance and decisions at key junctions.
---
Parallel in AI content:
Platforms like AiToEarn官网 and AiToEarn开源地址 show how full-stack AI strategies deliver hard-to-replicate advantages across industries — from automotive embodied intelligence to global content monetization ecosystems. In both, sustainable success comes from vision + decisive execution at the right time.
