Musk impersonation show. Has He Xiaopeng "survived until the final round"?

XPeng Motors still hasn't turned a profit, but its losses have further narrowed.

The financial report for the third quarter of 2025 shows that XPeng Motors' loss in that quarter was 380 million yuan, narrowing by nearly 80% year - on - year and 20.3% quarter - on - quarter.

XPeng Motors is at a critical turning point, and the speed of loss narrowing has exceeded expectations. However, there are still many challenges ahead for He Xiaopeng.

In the "imitation show" competition of domestic new energy vehicle companies for Elon Musk, He Xiaopeng is the one who has learned the most similar so far. However, since XPeng Motors has not yet made a profit, it may be too early to say that XPeng "has reached the final round".

A

Although the losses have narrowed and the delivery volume has soared, it is obvious that the market's expectations for XPeng are far higher than this report card.

The financial report shows that in the third quarter, XPeng Motors' total delivery volume was 116,007 vehicles, a year - on - year increase of 149.3%; the revenue was 20.38 billion yuan, a year - on - year increase of 101.8%. This growth rate is comparable to Tesla's early expansion stage. In the third quarter of 2018, Tesla's revenue jumped from $4 billion to $6.8 billion, a year - on - year increase of 129%, thanks to the large - scale delivery of the Model 3.

Judging from the monthly data, XPeng Motors' growth momentum is also strong. In August, it delivered 37,709 vehicles, a year - on - year increase of 169%; in September, it delivered about 41,600 vehicles, a year - on - year increase of 95%, breaking through the 40,000 mark and setting a new monthly delivery record for XPeng. In October, it continued to break through, delivering 42,013 vehicles, a year - on - year increase of 76%, setting a new historical high for monthly delivery volume and breaking through the 40,000 - vehicle mark for two consecutive months.

XPeng's gross profit margin in the third quarter reached 20.1%, which is the first time in the company's history to exceed 20%; the vehicle gross profit margin was 13.1%, an increase of 4.5 percentage points compared with 8.6% in the same period last year. This data is close to Tesla's in the third quarter of 2018, when Tesla's gross profit margin was 25.8% and the vehicle gross profit margin exceeded 20%. Moreover, in that quarter, the delivery volume of the Model 3 reached 56,065 vehicles, and the production man - hours decreased by 30% quarter - on - quarter.

However, the market's reaction is not as optimistic as the financial report data. On Monday, November 17, XPeng Motors' US stocks fell 6%, closing at $23.50.

Although the revenue in the third quarter reached 20.38 billion yuan, it was still slightly lower than the market expectation of 20.45 billion yuan; the non - GAAP loss per share was 0.43 yuan, also slightly worse than the expected 0.41 yuan.

More importantly, XPeng Motors' estimated delivery volume for the fourth quarter is between 125,000 and 132,000 vehicles, lower than the market estimate of 136,000 vehicles. It is expected that the revenue in the fourth quarter will be between 21.5 billion and 23 billion yuan, a year - on - year increase of 33.5% to 42.8%.

To make matters worse, the maximum estimated revenue of 23 billion yuan in the fourth quarter is far lower than the analysts' expectation of 25.36 billion yuan, with a gap of 15%.

What is even more worrying is the attitude of institutional investors. According to JPMorgan Chase's third - quarter position data, the bank reduced its holdings of 1.8239 million shares of XPeng's US stocks, and its position dropped sharply from the original 1.82 million shares to only 6,204 shares, a decrease of up to 99.7%. At the same time, NIO and Li Auto also suffered significant reductions in holdings.

However, at the conference call that night, He Xiaopeng was quite optimistic. He once again emphasized his previous view that XPeng is still expected to achieve profitability in the fourth quarter.

At the same time, He Xiaopeng also gave a grand plan for 2026.

In terms of products, XPeng plans to launch seven new models with "dual capabilities in one vehicle" in 2026. Among them, three super electric extended - range products will be launched in the first quarter, using longer pure - electric range and 5C ultra - fast charging technology; four new models will be launched throughout the year, covering important market segments. At the same time, XPeng plans to launch three new models for the overseas market, including small and medium - sized SUVs.

In the field of humanoid robots, He Xiaopeng announced that XPeng will achieve large - scale mass production of the high - level humanoid robot IRON by the end of 2026. This robot, which uses the "physical world model - driven" technology route, has been put into practical training at XPeng's Guangzhou factory, undertaking basic assembly tasks such as screwing.

"The long - term potential of the robot market will exceed that of the automotive market," He Xiaopeng said. "I expect that the annual sales volume of XPeng robots will exceed one million units by 2030."

This goal is close to the timeline of Musk's Optimus plan, but it is a bit more aggressive. According to Musk's plan, Tesla will produce thousands of Optimus robots for internal testing in 2025, increase the production volume to 50,000 - 100,000 units in 2026 and start selling them externally, and achieve a scale of one million units per year from 2029 - 2030.

XPeng's goal of mass - producing by the end of 2026 and reaching one million units by 2030 means that it needs to complete the leap from zero to one million within a shorter time frame.

B

Elon Musk, who once implied that He Xiaopeng was "copying", has started to praise this potential opponent across the ocean and even expressed a kind of mutual appreciation similar to "Among the heroes in the world, only you and I are worthy".

On November 5, XPeng Motors launched a new generation of IRON humanoid robot in Guangzhou. Musk then liked a video about XPeng's robot on social media and said: "Tesla and Chinese companies will dominate the market."

"I have great respect for Chinese competitors. There are many smart and hard - working people in China," Musk said.

Two days later, at Tesla's annual general meeting, Musk's trillion - dollar "compensation package" was approved. In a great mood, Musk "danced" with Tesla's humanoid robot Optimus V2.5 on the spot and announced that the production of Optimus V3, the third - generation Tesla humanoid robot, will start next year.

In addition, Musk claimed that Tesla's humanoid robot will be the largest - scale product in history, with an estimated market scale of billions of units. The ratio of private robots to industrial robots is about 1:3 to 1:5.

On the other hand, He Xiaopeng announced to the outside world that the global automotive market has a scale of $10 trillion and 90 million vehicles; the robot market has a potential scale of $20 trillion and 200 million units.

"From another perspective, I haven't thought about how many robots I can sell in a year after 10 years, but it will definitely exceed that of cars and exceed one million units," he said.

Both companies launched bipedal humanoid robots between 2022 and 2023. Now, with the product upgrades, their ambitions for embodied intelligence have become clearer.

In terms of technology, products, and vision, XPeng is quite similar to Tesla in many aspects and is regarded by many as Tesla's "student". And Musk initially joked and questioned, but now he is more willing to accept the other side.

In the early stages of new - energy vehicle startups, in order to differentiate themselves, each car company chose different development directions. For example, NIO chose battery swapping and services, while Li Auto focused on solving range anxiety and providing features like refrigerators, color TVs, and large sofas.

After some thinking, He Xiaopeng, the founder of XPeng Motors, established a strategy centered on "intelligence" and invested a lot of resources to build a full - stack self - developed system, covering everything from vehicle - end perception to cloud - based data operation.

This route coincides with Tesla's in many ways, and the results are reflected in XPeng's first mass - produced model, the G3.

At that time, the automatic parking function of the G3, with its high success rate and practicality, quickly established a good reputation among users and successfully labeled XPeng Motors as a "smart driving" brand.

What really enabled XPeng Motors to gain a foothold in the industry was the P7 model launched in 2020. Many people regarded this car as the "Chinese version of the Model 3" in terms of design and positioning.

The P7 was also the first domestic pure - electric vehicle equipped with the NGP (High - speed Navigation Assisted Driving System). In 2021, XPeng's annual sales volume was 98,000 vehicles, of which 60,000 were XPeng P7s.

However, XPeng didn't blindly follow Tesla.

In November 2020, XPeng Motors announced at the Guangzhou Auto Show that starting from the models produced in 2021, lidar would be added to the autonomous driving hardware system.

In contrast, Musk has publicly stated many times that relying on lidar is "stupid" and believes that it is an expensive and unnecessary sensor. To this day, Tesla still adheres to the pure - vision solution without lidar.

At first, Musk didn't take XPeng, this "student", seriously. And a dispute over the job - hopping of a former Tesla employee once led the two companies into a legal dispute.

At the end of 2018, after Cao Guangzhi, a senior engineer at Tesla (who was core involved in the development of the Autopilot system and had access to the source code), left the company, he quickly joined XPeng Motors as the head of the perception department. In March 2020, Tesla filed a lawsuit, claiming that he had stolen trade secrets. XPeng later denied this.

Soon after, Musk posted on social media, implying that "XPeng's autonomous driving technology seems to be 'highly similar' to Tesla's old code".

He Xiaopeng also issued a tough statement against Tesla, saying, "Starting next year (2021), in the field of autonomous driving in China, XPeng will beat Tesla so badly that it won't know where it is."

However, the lawsuit ended in a settlement between the two parties in 2023. Tesla withdrew the charges, and the specific settlement terms were not made public.

In the following two years, He Xiaopeng and Musk showed goodwill to each other from a distance, and their relationship eased.

In 2024, He Xiaopeng test - drove a Tesla car equipped with the latest FSD system in the United States and publicly shared his experience on social media. He said that he hoped to borrow a vehicle with the latest beta version of FSD for a comparative test with XPeng's XNGP.

In response to this public challenge, Musk's response was surprisingly friendly. He praised Chinese car manufacturers as the most competitive in the world.

In addition to cars, Tesla is promoting humanoid robots, and XPeng has also embarked on the same path.

In August 2021, Musk announced the humanoid robot plan, Tesla Bot, codenamed "Optimus", at the first AI Day. In October 2022, Tesla showed the prototype of Optimus at AI Day. This robot was already able to walk independently and carry objects. In December 2023, Tesla released the second - generation Optimus Gen2, which achieved significant improvements in flexibility and control ability.

XPeng acquired a robot company in December 2020, but it was not a humanoid robot company but a quadruped robot company. Immediately afterwards, it established Pengxing Intelligence and officially entered the robot track. There is still a big gap between quadruped robots and humanoid robots. Therefore, it wasn't until October 2023 that XPeng launched its first bipedal humanoid robot, the PX5.

One year later, at XPeng's AI Technology Day in 2024, the XPeng IRON was unveiled. Its hands are designed to be the same size as a human's, with 15 degrees of freedom and support for tactile feedback. He Xiaopeng emphasized the commercialization progress of IRON at the press conference, stating that it had been put into practical training at XPeng's Guangzhou factory, undertaking basic assembly tasks such as screwing. At that time, some production positions of the upcoming XPeng P7 + were already being operated by this robot.

But XPeng's mass - production plan at that time was not as grand as it is today: to achieve large - scale mass production of industrial - level L3 humanoid robots in 2026. The goal was to produce 10,000 units annually in the initial stage and reach a production capacity of 50,000 units in 2027.

There are a lot of commonalities in technology between smart cars and humanoid robots: the perception, decision - making, and control systems of autonomous driving can be directly transferred to the robot's "brain"; the hardware such as motors, batteries, and sensors can almost share the supply chain between the two; more importantly, the company's own factory provides the best testing and application scenarios. It is not surprising that Tesla and XPeng have both chosen the robot track from a technical - route perspective.

C

The reason why a company chooses a certain strategy often comes not only from catering to the market environment and customer needs but also from the DNA of the company's management.

Elon Musk made his first fortune in the Internet industry, and so did He Xiaopeng. This enables both of them to break out of the shackles of traditional automobile manufacturing and think about products and technologies from an Internet perspective.

In addition, Musk's management style is notoriously strict. He implements an extremely flat management structure, reduces middle - level managers, and communicates directly with front - line engineers.

He Xiaopeng learned Musk's flat management style but is more practical and gentle overall.

In 2022, the launch of the XPeng G9 was a failure. He Xiaopeng believed that the reason was internal decision - making and execution. Subsequently, he initiated a comprehensive and in - depth adjustment covering multiple fields. Only two of the 12 senior executives in the financial report remained, and there was an all - around integration of systems such as marketing, production, and R & D.

Later, He Xiaopeng said in Luo Yonghao's podcast that about 90% of the front - line departments had changed their leaders. This adjustment was a top - down all - around transformation aimed at completely reversing the company's decline and clearing the obstacles for the implementation of the new strategy.

After this adjustment, He Xiaopeng directly took over the reporting lines of all R & D, production, and supply - chain systems and even personally managed the styling team. At the same time, he established three styling front - end teams for creative competition, hoping to improve the design performance of future models in this way.

This management structure is the same as what Musk advocates, reducing intermediate levels to