Operator:

Hello, ladies and gentlemen. Thank you for standing by, and welcome to Pony AI Inc.'s Second Quarter 2025 Earnings Conference Call. [Operator Instructions] As a reminder, today's conference call is being recorded, and a webcast replay will be available on the company's Investor Relations website at ir.ponyai under the News and Events section. I will now turn the conference over to your host, George Shao, Head of Capital Markets and Investor Relations at Pony AI. Please go ahead, sir. George Shao: Thank you, operator, and hello, everyone. We appreciate you joining us today for Pony AI's Second Quarter 2025 Earnings Call. Earlier today, we issued a press release with our financial and operating results, which is available on our Investor Relations website. An earnings presentation, which we will refer to during this conference call can also be accessed and downloaded on our Investor Relations website. Joining me today on today's call are Dr. James Peng, Chairman of the Board and Chief Executive Officer; Dr. Tiancheng Lou , Chief Technology Officer; and Dr. Leo Wang, Chief Financial Officer of the company. They will provide prepared remarks followed by a Q&A session. Before we begin, please refer to the safe harbor statement in our earnings release, which applies to this call as we will be making forward-looking statements. Please also note that we will discuss non-GAAP measures today, which are more thoroughly explained and reconciled to the most comparable measures reported under GAAP in our earnings release available on our IR website and filings with the SEC. I will now hand it over to our Chairman and CEO, Dr. James Peng. Please go ahead. Jun Peng: Thank you, George. Hello, everyone. Thank you for joining our earnings call. We have actually had a very exciting quarter. Let me walk you all through 3 achievements that define our success this quarter. First, in June, we finished building assembly lines for our seventh generation robotaxi vehicles. In the last 2 months, we ramped up production capacity with more than 200 Gen-7 vehicles rolled off the assembly line. This puts us firmly on track to hit our target of 1,000-plus vehicles by year-end of 2025. We have operated Gen-7 robotaxis in all our 4 Tier 1 cities in China with over 2 million kilometers of open road autonomous driving. The path towards positive unit economics is also very clear as we made substantial reduction in key cost items such as remote assistance and vehicle insurance. Second, we continue to expand our robotaxi service areas and our user reach. The registered users surged by 136% year-over-year in Q2. In addition, we secured Shanghai's first fully driverless commercial license and expanded presence globally in cities such as Dubai, Seoul and Luxembourg. Third, our strong operational momentum has translated into outstanding financial results. On a year-over-year basis, our total revenues surged by a remarkable 76%. Robotaxi revenues have more than doubled with fare charging revenues growing more than 300%. Now let me share more details in these areas. As I mentioned previously, we positioned 2025 as a pivotal year of mass production. So first, let me talk about the mass production of Gen-7 robotaxi vehicles. The first 2 models of Gen-7, namely the Guangzhou Automobile Group, the GAC and the Beijing Automotive Industry Corporation, the BAIC models, have entered mass production in June and July, respectively. As of today, over 200 vehicles have been produced, which puts us well positioned to reach the 1,000-plus vehicle milestone by year- end. We have made significant improvements to our Gen-7 vehicles in terms of safety, cost, reliability and even appearance. The platform features a 100% automotive-grade autonomous driving kit designed for a 600,000-kilometer life cycle. All the sensors are highly integrated with the vehicle and can be preassembled. The total BOM cost has also reduced 70% compared to our previous generation. That is the cost dropped to even less than 1/3. Building on this foundation, we launched the operation of multiple Gen-7 robotaxi models in all four Tier 1 cities. This highlights how our autonomous driving systems can reliably adapt to different vehicle platforms and road environment. Since launching our Gen-7 vehicles have exceeded over 2 million kilometers of on-road autonomous driving mileage. The operational environment include extreme weather conditions such as tropical storms, high temperature and cold weather conditions, further validating our vehicle safety and reliability. So far, we have an impeccable safety track record the overall performance of Gen-7 Robotaxi also exceeded our previous generation. The newly designed in cabin interface provides a much better user experience. We have also made Robotaxi fleet operation much more efficient. Notably, we achieved reductions across key cost items such as energy, vehicle insurance, maintenance, remote assistance and ground support. There are two developments standing out in Q2. The first one is that we have sequentially improved the remote assistance to vehicle ratio, making us confident in achieving 1:30 ratio by year-end. This means that on remote assistant will be capable of monitoring 30 vehicles. The second highlight is that we have secured an 18% reduction in vehicle insurance costs compared with our prior renewal cycle. As I actually mentioned in the last quarter, our insurance premium was already at just half of the typical cost for traditional human- operated taxis. This further decrease in cost reflects the growing recognition by insurers of our safety track record. With mass production underway, licensed secured and service optimized, we are now well positioned to meet the rising demand from users. We continue to accelerate commercial deployment and expand our ecosystem at scale. Our commercial growth continued to gain momentum with global taxi service revenue up by 158% year-over-year. Particularly, we delivered more than 300% year-over-year growth in fair charging revenues. The impressive growth momentum was fueled by our accelerated deployment of global taxi across all four Tier 1 cities enabling us to expand into a broader range of use cases and boost higher user adoption. Additionally, our efforts to optimize the user experience significantly enhanced user engagement, reinforcing customers' recognition and trust in our Robotaxi services. To further accelerate this momentum, we continued our efforts in building a mobility ecosystem, in June, we formed a strategic partnership with Xihu Group, which is the largest taxi operator and fleet manager in Shenzhen. We plan to jointly deploy over 1,000 Robotaxis in Shenzhen in the coming years. This collaboration pairs our AI-powered technology and service expertise with Xihu's operational scale and deep local experience. Now let me switch gear towards expanding services to a broader user reach, we have seen strong user adoption in the second quarter with registered users surging by 136% year-over-year. Even with a larger user base, our user satisfaction rate remains well above 4.8 out of 5. Currently, we have secured permits in all four Tier 1 cities for our Gen-7 vehicles. As we continue to add more vehicles to our fleet, we expect to enter even larger scale operations for many more users to enjoy. Particularly, we are one of the first companies to receive a permit for fully driverless commercial Robotaxi service in Shanghai's Pudong District. Following the acquisition of this permit, we become the owning company to have commercial fully driverless operations in all four Tier 1 cities. This showcases both our technological maturity and solid operational capabilities. Recently, we also expanded the fare charging global taxi service from 15 hours per day to a full 24/7 coverage in Guangzhou and Shenzhen to better meet the rising user demand. This expanded operating window now captures late night and early morning trips. So now let me share the update on our global expansion. Our international playbook focused on entering strategic markets with hyper growth potential. To date, we have established presence in seven countries across China, East Asia, Europe and the Middle East. During Q2 this year, we made significant strides in Dubai, Seoul and Luxembourg. In the Middle East, we established our presence in Dubai, the largest and most populous city in UAE. We formed a strategic collaboration with the Dubai Roads and Transport Authority to integrate our autonomous driving technology into the city's transportation ecosystem. Initial trials will begin in late 2025, followed by fully driverless operations shortly after. We have also advanced our presence in South Korea by securing nationwide permits. We are conducting extensive on-road operations in Gangnam District, which is the heart of Seoul. We have navigated complex urban environments which showed reliability under challenging conditions such as winter snowfalls. In the second quarter, we began nighttime and early morning operations, which is a significant step towards 24/7 operations. Shifting to Europe, we have obtained a testing permits early this year in Luxembourg by partnering with Emile Weber, the Luxembourg's leading mobility and fleet service provider. We launched our road test in the city of Lenningen. Our goal is to use Luxembourg as a launchpad to serve future European expansions. In summary, we have laid a solid foundation for large-scale commercial Robotaxi operation. 2025 is the pivotal year of mass production for Pony AI, and we have already made it happen. With Gen-7 Robotaxi rolling off the production line and deployment across major cities globally. We are driving strongly towards positive unit economics and entering a new phase of multiple year accelerating growth. In the coming quarters, our main focus will be scaling up with a solid plan and great execution in place. I'm extremely excited about the future of Pony AI in revolutionizing the mobility industry. This concludes my remarks. Now I will hand it over to our CTO, Dr. Tiancheng Lou. Tiancheng, please go ahead. Dr. Tiancheng Lou: Thanks, James. So hello, everyone. This is in Tiancheng. I'm very proud of what we have achieved in the second quarter. We have started mass production of Gen-7 Robotaxi. This shows that we are scaling up our autonomous mobility technology. It also proves that we have constructed two important pillars for the Robotaxi industry, fully driverless and scale. These two pillars are the foundation of success of the Robotaxi business. Let me elaborate a little more. So fully driverless means no safety operator inside the car at all. This is a strong indicator that technology already meet safety standards. It also means the developer software and hardware system provides sufficient redundancy to handle extreme cases. So if something breaks, the car can still work well. After we get to fully driverless, the next pillar is scale. This refers to the number of fully driverless Robotaxis should be at least order of hundreds run on the road every day with a large enough of ODD. People can just open their ridesharing apps and call Robotaxi at their convenience. Based on these two standards, the development of L4 Robotaxi technology universally follows three conceptual stages. The first stage, cars still need a safety operator, either in the driver seat of passenger seats. The safety operator inside the vehicle has to step in to take over to mitigate any potential incidents. This incident may be caused by AI model errors, hardware failures or simply traffic scenarios that the vehicle cannot handle. Second stage, cars can run without any safety operation in the car, although it's still limited in a relative small scale, such as roughly a dozen of cars. At this stage, vehicles are equipped with redundant system, which can handle hardware failures. It also means the safety capability of the whole system reached roughly a similar level as a typical human driver. Moving to the third stage. Robotaxis can go fully driverless with hundreds of cars open to the public. They can also run 24/7 in all weather conditions, especially during rushovers. Once Robotaxis can operate without any safety operation in the car, it must reach a significant high level of safety to enable large-scale operations. Reason that statistically speaking, as the fleet size gets bigger and the cars drive more mileage every day, there is higher possibility for potential error to occur. As such, to scale up safely, the whole system has to meet much higher safety standards. As a result, the size of the fully driverless fleet is directly proportion to the high standard of safety and reliability. Only teams that truly trust their technology will run fully driverless service on a large scale. At this stage, the system can handle complex driving situations, and this is at least 10x safer than a human driver. That means users can easily access fully driverless Robotaxi service whenever they want, even through tough conditions like heavy rains or snowstorms. We will get a ride that is reliable and safe at all time. So we are proud that Pony AI have already achieved the third state in 2024 last year. With our current expanded operation in Shanghai is a good example. Let me elaborate. In July, we secured the newly issued fully driverless commercial license in Pudong District of Shanghai, and we offered the Robotaxi service to the public during World AI Conference, WAIC. We stood out in two ways. Number one, we were the only company with fully driverless and on-demand ridesharing service to the public. Number two, Shanghai was raining during the conference. We were the only company to remain operational in heavy rings. The fast speed at which we are scaling up fully driverless Robotaxi in Shanghai shows our strong generalization capability. This is powered by our high fidelity training environment and the evaluation system. Today, people in all four Tier 1 cities can call our fully driverless Robotaxi and get to anywhere they are going. Our latest Gen-7 Robotaxi platform powers a fully driverless scalable services. We also enforce regular engineering practice to enforce our safety-first principle. We have already produced over 200 Gen-7 Robotaxi and a proven they are ready for large-scale deployment. Our multiple Gen-7 models are equipped with multisensor switch and the new computing platform with four OX chips. This vehicle has achieved over 2 million kilometers on public roads across all four Tier 1 cities. The testing in diverse conditions shows reliability of our entire auto-driving specs, which must better integrate sensor setup and upgraded sensor clearing system the vehicle are much more reliable even in extreme weather and heavy traffic conditions. More importantly, the target of our 1,000 Robotaxi vehicles is just the start. As the service area expanded and the fleet gets denser, the network effect will kick in. This means that the user can reach more destinations and with less time for a ride. -- overall, a much better user experience. We have also enhanced the in-cabin experience for Gen-7 Robotaxis such as voice interactive features. All this will lead to accelerating multiyear growth trajectory. To wrap up, our achievement makes the second quarter reinforced our confidence in the path forward. Mass production is firmly on track with Guangzhou Auto and Beijing Auto Gen-7 vehicle production ramping up fast. We are continuously improving our operational efficiency and reducing the relevant costs. All this effort puts us in a strong position to scale up autonomous mobility rapidly. This concludes my prepared remarks. I will now pass the call to our CFO, Dr. Leo Wang, to review our financial results. Leo, please go ahead. Haojun Wang: Thank you, Tiancheng. Hello, everyone. This is Leo. I will be focusing on year-over-year comparisons for the second quarter unless otherwise specified. Q2 was another strong quarter marked by robust revenue growth and the significant progress in the mass production of our Gen-7 Robotaxi fleet, while also maintaining disciplined investment. In the second quarter, total revenues finished at USD 21.5 million, growing by 76% year-over-year. mainly driven by robust growth in both Robotaxis services and licensing and applications, a clear demonstration of our effective commercialization execution. Robotaxi service revenues reached USD 1.5 million, representing another quarter of rapid year-over-year growth of 158%. Both fare charging services and the project-based engineering solutions continued to expand strongly. In particular, fare charging revenues expanded by more than 300% year-over-year. The strong growth was driven by expanding user adoption and the demand in Tier 1 cities in China, as well as an increased fleet of Robotaxi vehicles. In addition, through ongoing optimization of our pricing and operational strategies across diverse user segments, we continued to enhance user engagement and the service efficiency. Such growth momentum underscores our commitment to establishing a scalable and recurring monetization model enhancing our long-term business visibility. Moving to Robotruck services. Revenues were USD 9.5 million as we proactively optimized our operation to focus on high-margin revenues. Robotruck services revenue were down by 10% year-over-year for the second quarter. Licensing and application revenues were USD 10.4 million, growing significantly by 902% year-over-year. We saw increasing orders and deliveries for autonomous domain controller sales, driven by both new and existing Robo delivery clients. Turning to gross margin, it improved year-over-year to 16.1%, with gross profit of USD 3.5 million in the second quarter. As mentioned in the last quarter, we continue to advance initiative aimed at reducing gross margin variability with a focused strategy on prioritizing high-margin revenues within Robotaxi and Robotruck Services. I also want to highlight that we made solid progress in optimizing Robotaxi unit economics such as remote assistance and vehicle insurance. Total operating expenses were USD 64.7 million, up by 75% year-over-year. The increase in share-based compensation expenses reflected the normalization of expense recognition following our IPO in November 2024, as vesting is no longer contingent of IPO completion. Excluding share-based compensation expenses, non-GAAP operating expenses were USD 57.5 million, up 59% year-over-year. The increase was primarily driven by increased investments in mass production, alongside employee expenses aimed at strengthening our R&D capacity for Gen-7 Robotaxi vehicles. Net loss for the second quarter was USD 53.3 million, an increase from USD 30.9 million in the same period last year. Non-GAAP net loss was USD 46.1 million compared with USD 30.3 million in the same period of last year, primarily reflecting increased investments in mass production and R&D employee expenses for concurrent development of Gen-7 vehicle models. As we drive towards large-scale Robotaxi mass production and deployment, our disciplined investments remain a top priority as we enter this critical scaling phase. Turning to the balance sheet and cash flow. Our combined cash and cash equivalents, restricted cash, short-term investments and the long-term debt instrument for wealth management was USD 747.7 million as of June 30, 2025. The quarterly financing cash inflow increased by USD 33.1 million mainly due to employee share sales following the expiration of the lockup period, resulting in funds collected on behalf of employees for future distribution. With the imminent scaling up and the commercial deployment, we believe our current cash reserves are well positioned to support our operational needs. In the same time, we remain proactive in exploring additional opportunities to ensure long-term financial resilience. Looking ahead, with disciplined investment in place and mass production underway, we are very on track to reach 1,000 Robotaxi fleet size target, already with over 200 contributed by the Gen-7 Robotaxi. With a competitive cost structure for the latest Gen-7 vehicles and a well-executed go-to-market strategy, we have laid a solid foundation for large-scale commercialization moving forward. I will now turn the call over to the operator and begin our Q&A session. Thank you. Operator: Thank you. We will now begin the question-and-answer session. [Operator Instructions] Our first question is from Ming-Hsun Lee, Bank of America. Ming-Hsun Lee: Congratulations on the solid progress of your mass production. So I only have one question. Could you share more thoughts on your production plan throughout the second half of 2025? Jun Peng: This is James. I'll take this call. As I mentioned in my prepared remarks today, we are firmly on track to reach our 2025 production target. In less than 2 months, we have already ramped up production with over 200 Gen-7 Robotaxi vehicles it showed a clear accelerating momentum. As of now, we are very confident in surpassing a fleet size of 1,000 plus by year-end. The key actually to achieve the Robotaxi mass production lies in the ramping up efficiency, which involves comprehensive process such as equipment calibration and also the staff training. These steps ensure a steady increase in production volume while upholding the highest safety standards. As of now, we have already reached a scalable production cycle. In addition, the key components are already secured through an agile and holistic sourcing strategy. This not only ensures sufficient inventory to support production pace but also helps to maintain the BOM cost stability. Currently, we have completed the retooling of assembly lines for both the BAIC and the DAC Gen-7 Robotaxi models. Both of these models have entered the SOP phase. That is the start of production phase with steady production rate, reflecting our strong execution. These vehicles have begun on-road testing. And so far, we have impeccable safety track record, with more than 2 million kilometers of operation, and the results confirm they met all our stringent safety requirements. Now let's also look at the unit economics. The BOM cost of over Gen-7 Robotaxi has been reduced by 70% compared with the previous generation. During Q2, we have also made major improvements in Robotaxi fleet operations such as enhanced remote assistance efficiency and decrease in the vehicle insurance. This is further reinforced by our proven go-to-market strategy as we have seen a strong growth in fare charging revenues. So with all these improvements I'm actually now very confident in reaching positive unit economics for our Gen-7 Robotaxi vehicles. With this, back to the operator. Operator: Next question is from Bin Wang, Deutsche Bank. . Bin Wang: Your robotaxi revenue growth was very strong once again this quarter. Can you elaborate the key drivers behind the solid growth? For example, what's the fleet size at the beginning of the quarter and the end of the second quarter? Haojun Wang: Yes, this is Leo. I'll take this question. As I mentioned earlier, our Robotaxi service revenue rose by 158% year-over-year. powered by more than 300% expansion in fare charging revenues. This growth actually stems from expanding user adoption and the demand in Tier 1 cities in China. . Coupled with an increased deployed fleet of Robotaxis. Through ongoing optimization to our pricing and also operation strategies across diverse user segment. We have greatly improved the user engagement and service efficiency. Our current user satisfaction rate is well above 4.8 out of 5. Additionally, we are currently operating across 2,000 square kilometers in Tier 1 cities in China. That's more than 20x the size of City of San Francisco. This provides us ample market space to deploy a larger- scale Robotaxi fleet. So looking ahead, increasing vehicle density in these regions remains our top priority. We expect notable improvements in vehicle accessibility and user experience, which, in turn, will incentivize higher demand. By continuously scaling up production and deployment, we are well positioned to improve vehicle utilization and user experience. We believe these factors will be the key drivers of a scalable and a recurring monetization model that strengthening our business visibility over the long term. Now I'll get back to the operator. Operator: Next question is from Ting Song, Goldman Sachs. Ting Song: I have one question. Currently, China's Ministry of Public Security commented on the intelligent driving systems and autonomous driving. What does this mean for the full Robotaxi industry? And are there any changes in the government's attitude on the industry? Dr. Tiancheng Lou: This is Tiancheng. I will take this question. The recent comment you mentioned actually reinforced the distinguishing between L2 driving assistant and L4 autonomous driving. The government [ commented ] that L2 and L4 are two completely different products. Current L2 driving assistant intelligent driving system have not yet achieved fully autonomy that L4 can achieve. And on the L2, the driver remains ultimately responsible for vehicle operation. This is aligned with the MIT comment early in April this year. They offset a clear boundary that L2 is driving assistance system and L4 is autonomous driving. So this is actually beneficial for us. It helps the public clearly understand the distinguishing between L2 and L4 systems, and reduces the risk of user mistakenly treating L2 as for L4 autonomous driving. In addition, the difference in responsibility will result in different safety standards. The L2 systems liability lies with the drivers. In contrast, the L4 system have full responsibility, resulting in a safety first principle and rigorous engineering practice. Given above, Pony AI is a company dedicated to L4 solutions, achieving L4 autonomous driving requires very high safety standards. It will also needed a multilayer software and hardware architecture that provide redundancy in the event of system component failures. It must include critical components of sensors, computing, controls and power systems. In China, to obtain fully driverless permit, redundancy system must be tested in many complicated situations such as handling vehicles coming out of blind spot. That's why most mass produced vehicles out there need to add extra redundancy to meet the basic requirement. With that, back to the operator. Operator: Next question is from Xuyang Li, [indiscernible] Securities. Unidentified Analyst: So first of all, Congrats on recent business development expansion. So my question is, as we look at new market expansion, what are the key technical requirements to enter new geographies? And how confident are you in your system's ability to adapt quickly and safely to different environments? . Dr. Tiancheng Lou: This is Tiancheng here. The core of L4 auto driving is all about the ability to handle corner cases and extreme cases. In different geographics, this type of scenarios are actually similar. For example, a pedestrian stepping into the road, a vehicle suddenly changing lanes or there's unexpected road construction. These situations are fundamentally the same even if their frequency can be varied by location. So our software system have achieved fully driverless and scalable and 20% operation across different environments. This key enabler is our point word model, which we deploy consistently across cities and countries. It enables us to generate numerous challenges scenarios and replicate existing scenarios with many variations. This significantly enhanced AI driving model's ability to handle a wide range of challenging situations. While entering a new area, we can quickly launch road testing and validation. Additionally, turning of the AI model is not needed. So for example, our current operation at WAIC in Shanghai have already proved that our system is robust, adaptable and built for rapid replication. This resilience is also enhanced by dealing with complex urban environment. This gives us a solid foundation for efficient and reliable expansion into large ODD and scalable operations. So with this, back to the operator. . Operator: Next question is from Xinyu Fang UBS. Xinyu Fang: Congrats on rapid growth in Robotaxi services this quarter. My question is that we've seen news reports about potential Hong Kong IPO plans of the company. Do you have any timetable that you can share for now for such IPO, please? Thank you. . Haojun Wang: Yes, this is Leo. I'll take this question. First of all, I would say we do not comment on market speculation. Additionally, we actually have been closely monitoring the market conditions. In line with our strategic priorities of scaling up, we remain focused on executing our go-to-market strategy. Meanwhile, we will actively explore more capital market opportunities to maximize shareholders' value. I'll get back to the operator. Operator: Next question is from Eveline Zhang from Daiwa. Evelyn Zhang: So my question is about your overseas market expansion. Could you please elaborate more on your future plan to expand your Robotaxi business in the overseas market? And as you mentioned earlier, you have deployed Robotaxi in Middle East and in South Korea and Europe. And could you please share your insights on how overseas regulation framework looks like? For example, your Robotaxi in these 3 areas are still in the road testing stage or in the commercial service stage right now? Jun Peng: This is James. Let me begin with an overview of our global expansion strategy. As you can see that our current global expansion actually focused on markets with hyper growth potential. That is, we are focused on markets with strong mobility demand well- developed infrastructure and supportive regulatory environments. So we have built our process through deep collaboration with the local partners and policymakers. So given our proven ecosystem, technology expertise and large-scale operation experience in China, I think we are very much well positioned to rapidly replicate of our success to the international markets. As I mentioned in my prepared remarks, during Q2, we entered Dubai as a new market, advanced our operations in South Korea. We also strengthened local partnership in Luxembourg, where road tests have now started. Every step forward reflects our core belief autonomous mobility everywhere. In terms of the our operational and some details of our global expansion. I'll hand it over to Leo, who will give you further perspective. Haojun Wang: Thank you, James, and this is Leo. So from an operation perspective, scaling to hundreds of commercial fully driverless Robotaxis requires strong local policy support, technology cost competitiveness and user adoption. As of now, a few countries have achieved this level of readiness and deployment except for the U.S. and China. Other than these two countries, most countries only have the regulation allowing the testing with a safety driver. And most of these commercialization remain limited project-based or one-off pilot program such as vehicle sales as of now. We believe it still takes time for a sustainable far-charging business model to mature in terms of both monetization and the unit cost front in this region. I would also emphasize that scaling full driverless Robotaxis fleet demand a fundamentally higher level of safety and operation rigor, and it becomes increasingly critical as the operation expanding. This leap from dozens to hundreds is a big jump in complexity. So building this operational capability is very essential. Only then can we validate the model and begin replicating it in other cities or countries. With that, I'll get back to the operator. Operator: As there are no further questions, I'd like to turn the conference back over to the company for closing remarks. George Shao: Yes. Thank you once again for joining us today. If you have any questions, please feel free to contact our IR team. We look forward to speaking with you in the next quarter. Operator: This concludes today's conference call. You may now disconnect your lines.