Lepton's vehicles barely break even; all profit comes from proprietary parts
Full-stack self-development dropped component costs below supplier rates. Lepton passed savings to customers instead of shareholders. Vehicles break even. The company survives on parts-margin alone.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Seeing millionaires around you is better than any business school
Zhu Jiangming grew up in Yiwu in the 1980s when farmers in the small-goods market were getting rich overnight. The wealth erupting around him wasn't demoralizing—it was proof of concept. He realized later that entrepreneurship isn't a decision people make in the abstract; it's something people do because the environment proves it works. In places where nobody starts companies, 1-in-10,000 will try. In boom towns, 1-in-10 will jump. He attributes his entire founder career to proximity and observation, not genius.
— Zhu JiangmingThe invisible market gap became a billion-dollar product
In the early 1990s, Zhu was working in remote surveillance. At a trade show in Beijing, he noticed the market had no real solution—vendors were just stitching together modules. Banks used cassette recorders for surveillance, producing eight hours of grainy footage. He saw the gap: someone needed to design a purpose-built digital recorder with its own compression chip and embedded CPU. He built it. Dahua's first DVR didn't open a market segment; it pulled the entire surveillance industry from millions per year in custom work into tens of billions in device sales.
— Zhu JiangmingIgnorance was essential—knowing the true cost would have stopped the venture
Before founding Lepton, Zhu's team had almost no contact with automotive industry veterans. They hired pure engineers and designed the car from first principles. He admits now: if he'd understood upfront that serious car manufacturing requires 200 billion RMB or more, Lepton never would have launched. He calls it ‘the luck of ignorance’—both curse and blessing. If entrepreneurs actually calculated the failure rate (under 1%) and capital required, almost nobody would start. Lepton succeeded partly because Zhu didn't know enough to be afraid.
— Zhu JiangmingA coupe designed in a villa weekend nearly bankrupted the company
Lepton's first car, the S01, was a two-door sports coupe. The decision came from a young investor who loved the styling. A few tech founders—none of whom knew consumer car markets—met at a villa in Qiandao Lake and decided on impulse to build it. Zhu reflects that a startup with zero brand recognition has no room for cars that live on styling alone. An EV without marketing overhead needs no explanation to sell. The S01 at 110,000 RMB was positioned wrong; if Lepton had discounted to 70,000 RMB like rivals did, the company wouldn't have survived. The lesson stuck: every car after S01 was designed with ruthless discipline.
— Zhu JiangmingNearly killing the bestseller-to-be saved the company
In spring 2020, Lepton was stretched for cash and nearly shelved the C11 project (then called the 411). Zhu decided to bet everything: design it to GLC-class Mercedes specs, target 200,000 RMB, and use pre-orders with tiered discounts to create scarcity. C11 launch exceeded expectations by orders of magnitude and became Lepton's true inflection point. S01 had been a learning disaster; T03 was just in-between. C11 proved that Lepton's design instinct was real.
— Zhu JiangmingVehicles break even; all profit hides in self-made parts
Vertical integration lets Lepton own the 10% margin that suppliers would normally capture. Instead of pricing high to keep that spread, Lepton prices aggressively and passes savings to buyers. Result: vehicles operate at near-zero profit, and the company survives on margins baked into its own semiconductors and components. Zhu compares Lepton to Uniqlo: thirty-yuan shirts that don't feel cheap even to people like him. The brand doesn't earn from pricing power; it earns from being trustworthy enough that quality-obsessed customers don't hesitate. All the money is in the parts.
— Zhu JiangmingWestern OEMs now reverse-engineer the partnership, not the other way around
In 2023, Stellantis invested 1.5 billion EUR and formed Lepton International, a 49%-51% joint venture to sell Lepton tech into markets outside China. This inverts decades of automotive playbook: Western companies exported technology to China in exchange for market access. Here, Lepton exports technology; Stellantis brings overseas channels, brand credit, and financial infrastructure. Zhu sees this as inevitable: after 40 years of industrial accumulation, China owns the final frontier—new-energy vehicles. It happened in appliances and phones; now it's happening in cars. The tech is Chinese-led; partnerships just become distribution.
— Zhu JiangmingRefusing to retaliate against attacks makes personal attacks pointless
Lepton publicly committed to never deploying fake accounts, badmouthing competitors, or spreading dirt. Zhu's logic: mudslinging is like two kids fighting—once you hit back, the conflict spirals. But if you get hit and don't retaliate, the other side has no sport and quits. He also believes illegal smear campaigns should trigger legal action, not silence—a hybrid stance: don't engage in mudslinging, but don't absorb legal violations either. So far, the commitment has held, freeing Lepton from spending energy on reputation defense.
— Zhu JiangmingIn their own words · checked verbatim
China still has four hundred million people riding e-bikes. If we convert them into four-wheel EV users, that's the whole opportunity.
中国还有四亿人口 就是骑电驴 是吧 把这些人口 能够转换成这个 这个四轮电动车的用户
Zhu Jiangming56:37
If we'd known upfront it would take two hundred billion, we probably wouldn't have started. We definitely wouldn't have. In reality, two hundred billion is the bare minimum.
所以当时要知道烧200亿 可能就不启动了 那肯定不会干 那实际上200亿是最起码的
Zhu Jiangming1:01:41
This can't be done without real skin in the game. Time has proven it: the founder has to actually put their own money in for it to work.
所以这个事本来没点加点是干不了的 所以就是说时间也证明 只有这个老板自己真正投钱了 这个才能够做成功
Luo Yonghao1:40:18
I keep comparing Lepton to Uniqlo. Uniqlo has twenty-or-thirty-yuan shirts, but even Luo or I don't feel cheap wearing them.
我一直在解救说 我说我们说这个领跑 是这个车界的优衣库 那优衣库 它有这个二三十块钱的衣服 但是这个这个 哪怕就是罗老师或者我都去买就是也不觉得low
Zhu Jiangming1:52:37
We only make money on parts margins—and that's exactly how we operate.
就是实际上就是我们只赚这个零部件的差价 这个确实我们是这么在干的
Zhu Jiangming2:01:40
If you don't have total confidence in product quality, you burden yourself. We can say this because we trust what we build.
这个因为如果说对产品质量就是说没有足够信心的话 那只会这个这个对你的公司带来更大的负担 我们只所以能敢这么说就是对我们产品的充分的自信
Zhu Jiangming2:17:02
Mudslinging is like two kids fighting—if you refuse to retaliate, there's no real conflict.
这个我的一个说法就是说 不然的话就是说 就其实这个 这个这个干这些脏活 就像这个两个小孩子打架一样
Zhu Jiangming3:04:00
Figures
| Estimated minimum capital to launch an EV maker | 200 billion RMB | 1:01:41 |
| Dahua market-cap peak | 900 billion RMB | 53:30 |
| Dahua annual profit range | 30-40 billion RMB | 38:27 |
| Stellantis investment in Lepton | 1.5 billion EUR | 2:23:13 |
| Lepton International joint venture ownership | Lepton 49% / Stellantis 51% | 2:29:16 |
| Zhu Jiangming personal stock buyback (annual) | 1.9 billion HKD | 2:25:14 |
| Lepton annual sales target | 1 million units | 2:11:57 |
Glossary
- DVR
- Digital video recorder that compresses and stores surveillance footage directly to hard disk; breakthrough product that scaled the security industry from bespoke systems to mass-market devices.
- IGBT
- Insulated-gate bipolar transistor; core power semiconductor in EV drivetrains that Lepton manufactures in-house to control costs and differentiate performance.
- NPS
- Net Promoter Score; metric measuring customer likelihood to recommend a product; indicates customer loyalty and word-of-mouth strength.
How to listen
Founders and investors curious about EV startup unit economics and full-stack R&D strategy.
First ~17 minutes of childhood and university background; low-density material, skip to the founding story.