The Past Story of Didi's Automobile Manufacturing: Regrets, Fulfillment and New Life
As Robotaxi stands on the eve of large-scale commercial deployment, hardware capability has become the third core threshold following AI and operation. Few people know that Didi, a mobility company that started out with ride-hailing services, once independently developed and manufactured a vehicle with full practical investment.
In 2021, Didi launched an in-house full-vehicle R&D project codenamed "Da Vinci", with its product named MONA. Two and a half years later, trapped by qualification and capital pressure, the vehicle failed to be launched under Didi's own brand, and was sold in its entirety to XPeng Motors, which has since ranked first in sales volume for 24 consecutive months.
This vehicle manufacturing attempt that did not end under Didi's own name left the company with accumulated experience in full-vehicle definition, engineering verification, and software and hardware integration. No one knows whether these capabilities can be translated into advantages in the Robotaxi era. But one thing is certain: the real competition has already started ahead of schedule.
In 2026, a mobility revolution that has been brewing for a decade is approaching its breakout point.
Waymo has launched Robotaxi services in 15 cities across the United States, Tesla Cybercab has hit the roads, and multiple Chinese cities have accelerated the commercialization of autonomous driving. The superposition of multiple signals means that Robotaxi is about to cross the inflection point and enter a new stage of large-scale commercial deployment.
A concurrent trend is that the importance of hardware to Robotaxi is rising, and multiple players have successively launched new models tailored for Robotaxi scenarios. The competition threshold for autonomous driving is further raised: the three capabilities of AI, operation and hardware are all indispensable, and there are very few companies that possess all of these capabilities at the same time.
Amid all the hype, few people connect Didi with "vehicle manufacturing" — compared with autonomous driving algorithms and operation systems, Didi's full-vehicle R&D capability has barely been fully recognized by the outside world in the past.
Even fewer people remember that this mobility platform once quietly launched a full-vehicle project codenamed "Da Vinci", whose team once expanded to nearly 2,000 people, who poured all their understanding of mobility scenarios into a vehicle codenamed MONA. But the vehicle was never released under Didi's name in the end.
In August 2023, trapped by qualification and capital pressure, Didi sold its smart vehicle development related business to XPeng Motors. On August 27, 2024, MONA M03 was launched and became an instant hit, and it is still the pillar that accounts for half of XPeng Motors' total sales so far.
After the Da Vinci project, the core technical team was merged into Didi's autonomous driving system, retaining years of accumulated experience in vehicle styling design, product definition, engineering implementation, software and hardware integration, and system engineering capabilities.
Recently, Snow Leopard Finance interviewed a number of industry insiders, pieced together and restored the little-known past of Didi's vehicle manufacturing, and tried to re-understand: why Didi made vehicles, why it gave up, and how this experience will affect Didi today.
Stepping into a Storm
The launch of the Da Vinci project happened to fall in a delicate time window.
2021 was the most distinct watershed for China's new energy vehicle industry. On one hand, the illusion of internet players making vehicles collapsed: a number of new energy vehicle enterprises that were once prosperous fell into difficulties one after another, even failing to cross the life-and-death line from 0 to 1, and the industry began to bid farewell to unregulated growth. On the other hand, it was the dawn before the arrival of a new era: Xiaomi, Baidu, Huawei, a group of cross-sector giants from the technology and internet industries entered the market in a high-profile manner, trying to restructure the competition landscape through the approach of "dimension reduction strike".
Didi's decision to make vehicles seemed extremely bold, even overly radical at that time.
At that time, although Didi was the absolute industry leader in the ride-hailing track, it faced fierce competition, and its attempts at diversified expansion encountered setbacks for several times. The autonomous driving technology that it had invested in for 5 years had not yet seen the light of day, its community group-buying business Chengxin Youxian burned through tens of billions of funds, and the difficulty of vehicle manufacturing was raised to a higher level. Didi's disadvantages were obvious: it had no vehicle manufacturing genes, no accumulated supply chain experience, and pending qualification licenses.
This seemed like a high-stakes gamble. But on the other side of the coin, Didi's advantages were equally clear: a mature software ecosystem, a huge pool of drivers and users, and a deep understanding of mobility scenarios, which are unique assets that other companies cannot easily replicate.
At that time, the industry had no answer as to whether these advantages could be translated into full-vehicle competitiveness. Entering full-vehicle R&D and manufacturing across sectors was in itself a high-risk experiment. Both the capital market and the industrial circle had obvious divergences on whether latecomers could still break through the encirclement.
Before the Da Vinci project, Didi had already had a rehearsal close to "vehicle manufacturing".
In 2017, Didi Automotive Innovation Center was established. In 2020, D1, a customized ride-hailing vehicle developed by Didi in cooperation with BYD, was launched. This was the first model designed exclusively for ride-hailing scenarios, solving some pain points in the ride-hailing scenario, such as battery range, connection between the vehicle and Didi's platform, single-sided sliding door, and optimized rear space.
However, the D1 project also made Didi deeply realize that only doing product definition and software design is far from enough, and it must tackle the hard nut of software and hardware integration — Didi decided to try to do it on its own.
What kind of vehicle to build? Didi spent half a year deliberating repeatedly to figure out this question.
The first question was whether to target the B-end or the C-end, which was a strategic divergence related to top-level product positioning. Various different voices emerged internally, and Didi finally came to the judgment that making a mass product for the C-end was a more important and correct choice.
The logic is simple: if you only target the B-end, it is like swimming in a pool that never grows up, and you will never have the imagination space of the ocean. In the field of household vehicles, sedans are the mainstream track with the highest popularity and the most likely to produce hit products.
At the same time, if a mass-market model is polished to the extreme, it can not only attract a huge number of C-end users, but its mature mass production system and cost control capabilities will also naturally feed back to the shared mobility scenario, realizing experience upgrade and cost optimization for ride-hailing services.
Finally, Didi positioned the product in the A-segment vehicle market, aiming to design a high-appearance, truly smart, cost-effective electric vehicle priced at around 100,000 to 150,000 yuan for young users, codenamed MONA.
The composition of the Da Vinci project team itself is an annotation of how Didi understands vehicle manufacturing: members need to understand vehicles, mobility, and the internet.
Based on the D1 project, Didi quickly expanded its team to nearly 2,000 people. Among them were not only elite members who had achieved good results in Didi's other businesses, but also "veterans" from the automotive industry.
A large number of mobility scenarios anchored the product direction, and the mainstream A-segment vehicle market provided a large enough battlefield. At this point, the foundation of the Da Vinci project was laid.
The direction was set, the team was formed, and the next task was to fight a real tough battle.
Climbing to the Peak, Finding No Road Ahead
For the people participating in the Da Vinci project, there was a clear consensus: the A-segment pure electric sedan is the market with the most sufficient competition and the most difficult to achieve differentiation. If the product is only "not bad", it will be difficult to gain user recognition, which means there is no chance of winning.
Precisely because the competition is too sufficient, even if the product strength is leading, whether it can finally become a hit is still affected by multiple factors such as brand, channels, price and marketing.
In traditional automakers, flagship models often get more resources and time allocation, and few people will put too much effort into A-segment vehicles. But MONA was different, as Didi's first model, it got the company's best resources and the full dedication of the entire team.
The name Da Vinci also hides Didi's expectations for the project. It implies the integration of technology and art, and also represents the creation process from 0 to 1. The legend of this Renaissance master lies precisely in the fact that he never separated engineering from aesthetics. The naming of the first product MONA comes from Leonardo da Vinci's most famous work, Mona Lisa.
Didi internally put forward a standard that a good-looking vehicle should have 3S: Simple, Smart, Sophisticated.
"To make most people like you, the style is most likely to be simple, because it is more inclusive, just like Apple's products. Everyone thinks a plain white T-shirt looks good, but if you print patterns on it, some people will like it and some won't," a member of the Da Vinci project team summarized, "The so-called sense of high quality comes more from the body proportion, posture, and the strong contrast between the complexity and simplicity of design elements."
These design concepts are not unique to Didi, but the common pursuit of the new energy vehicle industry in recent years. The greater challenge is how to implement these concepts under limited costs.
In order to find this balance, the design team of 50 to 60 people spent more than a year deliberating repeatedly, and the solutions went through multiple rounds of screening from 8 to 3, 3 to 2, and 2 to 1.
The really difficult part took place in the engineering phase.
In order to control the total height of the vehicle at 1.44 meters, the space reserved for the motor and battery was continuously compressed. To this end, the engineering technical team used a 118mm-high battery pack and a 1.2-square-meter panoramic sunroof. The battery pack became smaller, but the cruising range could not be reduced, so the team made great efforts to make the vehicle's drag coefficient the world's lowest at 0.194, to make up for the cruising range pressure caused by the limited battery capacity. With a 52kWh battery, the vehicle can achieve a cruising range of about 500 kilometers. The improvement of energy efficiency also saved 20,000 yuan in manufacturing costs, and the saved money was used for higher-level intelligent configurations.
Similar trade-offs ran through the entire development process.
Also in order to control costs, the Da Vinci team put a lot of effort into chassis tuning. "Others tune 5 or 6 rounds, we tune more than 20 rounds, and prepared more than 200 sets of elastic components," an insider revealed.
This unconventional "excessive investment" is not common on A-segment vehicles. However, no one could be sure whether this investment would eventually translate into market success until the product was launched.
After the prototype vehicle was completed, Didi internally organized three rounds of blind tests: randomly selected 200 users, placed 5 vehicles of the same class together, covered all brands, prices and configurations, and let them choose their favorite one by intuition. In the three rounds of tests, the selection rate of MONA remained between 67% and 70%.
However, this kind of internal blind test is more like a verification tool in the product development process, and the test environment also excludes key variables such as brand awareness and price sensitivity in real car purchase decisions, so there is still a gap between it and real market purchase behavior. But for the Da Vinci team at that time, this was at least a positive signal in product definition.
In any case, in this two-year sprint, the Da Vinci project completed tasks that traditional automakers would take three or four years or even longer to finish.
In July 2023, the design and R&D of MONA were all completed, and it entered the production preparation stage. But after climbing to the peak, they suddenly found that there was no road ahead.
The dream hit the south wall of reality. This vehicle was never launched under Didi's name in the end.
Separation, a Reluctant but Correct Choice
The subsequent story is the version known to the outside world.
In August 2023, Didi sold its smart vehicle development related business, and XPeng Motors took over with shares with a maximum consideration of about 5.4 billion yuan, and the new project continued to use the name "MONA". But XPeng gave it a brand new explanation: Made of New AI.
According to the disclosure of the transaction announcement, in addition to the share consideration at the two nodes of the first-phase asset delivery and SOP mass production, if the MONA model reaches the established sales target for two consecutive years, Didi's shareholding in XPeng Motors can rise from 3.25% to 5%. That is to say, the market performance of the model directly determines the final value of this asset, and the two sides have changed from a buyer-seller relationship to a community with shared interests.
It is not hard to imagine that this was undoubtedly a bold decision for XPeng. At that time, XPeng Motors was in a product strategic dilemma where the popularity of its old main models declined, new models were not yet ready, and there was a gap in the mid-end price band. In order to quickly reverse the situation of slowing sales and operating pressure, and re-stabilize its position in the top tier of new vehicle-making forces, it had to quickly launch a big hit product that could quickly open up the market.
But for an automaker that is good at self-research and has always emphasized full-stack self-research, buying a model developed by an external team is in itself a challenge to the inside. There must have been trade-offs between continuing to wait in the original self-research rhythm, or accepting a mature external solution to achieve rapid mass production and win time.
And MONA is precisely positioned in the largest entry-level high-volume track, filling the gap in XPeng Motors' product matrix. More importantly, developing an A-segment sedan from scratch has a long cycle, high investment, and time-consuming supply chain running-in, which is likely to miss the short window period in the fiercely competitive market environment. Didi has already completed the full process of full-vehicle R&D, engineering verification, and cost calibration, which can be quickly implemented after being taken over, and its value is exactly matched.
According to a person close to the transaction, what really dispelled XPeng's concerns was the obvious high completion level of MONA. After personally leading the team to test ride MONA, He Xiaopeng, Chairman and CEO of XPeng Motors, left a professional team to start technical due diligence. Soon, He Xiaopeng unified the internal consensus and decided to introduce a highly completed external model through equity swap.
At the exchange meeting after the transaction announcement was released in August 2023, He Xiaopeng also admitted that MONA is a highly completed pre-mass-production A-segment smart electric sedan, "a very attractive asset".
After taking over, XPeng Motors made some adjustments to the product, rolling out more comfort configurations to the base version, and replacing the supplier's solution with its self-developed intelligent driving system for the high-end version. The MONA seen in the market today is not a simple copy of Didi's version, but a product that has undergone redefinition and engineering optimization by XPeng Motors.
How good is the product strength of the adjusted MONA? The market has given the most intuitive feedback.
In April 2024, XPeng Motors officially announced the MONA series for the first time; in August of the same year, the officially launched MONA M03 became an instant hit, with more than 30,000 paid pre-orders in two days, and its monthly sales have steadily exceeded 10,000 units since the next month, becoming the core high-volume pillar of the company.
By June 2026, MONA M03 has ranked first in the sales of 100,000 to 200,000 yuan pure electric sedans for 24 consecutive months. Among them, in April and May of this year, its monthly sales exceeded all fuel sedans for two consecutive months.
The fact that MONA can finally be launched and continue to sell well is inseparable from XPeng Motors' continuous investment in intelligent driving, supply chain, manufacturing and brand marketing in the follow-up period.
However, behind this beautiful breakthrough battle, the full cooperation of both sides is indispensable.
Didi's efforts in product design, coupled with XPeng Motors' investment in intelligent driving, enabled MONA to deploy high-level intelligent driving functions to models at the 150,000-yuan price band, becoming "the first car loved by young people".
On July 2 this year, the first SUV of the MONA series — XPeng MONA L03 was officially unveiled and opened for pre-sales, with a global layout and exports to 64 countries and regions. MONA has once again become an important product driving XPeng Motors to accelerate its pace of going global.
Looking back, Didi's choice to hand over MONA to XPeng Motors is a win-win choice that meets each other's needs: XPeng quickly recovered cash flow through this