WeRide Inc (WRD) 2025 Q3 法說會逐字稿

Good morning, and good evening, ladies and gentlemen. Thank you for standing by. Welcome to WeRide's third quarter 2025 earnings conference call. (Operator Instructions) Please note that today's event is being recorded. The company's unaudited financial and operating results were released by the Newswire earlier today and are currently available online. Joining us today are WeRide's Founder, Chairman and CEO, Dr. Tony Han; and CFO and Head of International, Ms. Jennifer Li.

女士們、先生們，早安，晚上好。感謝您的耐心等待。歡迎參加 WeRide 2025 年第三季財報電話會議。（操作說明）請注意，今天的活動正在錄製中。該公司未經審計的財務和營運業績已於今日稍早由新聞專線發布，目前可在網路上查閱。今天與我們一同出席的有 WeRide 的創辦人、董事長兼執行長韓東尼博士，以及財務長兼國際業務負責人李女士。

Before we continue, I would like to refer you to the Safe Harbor statement in the company's earnings press release, which also applies to this call as today's call will include forward-looking statements, including WeRide's strategies and future plans. These forward-looking statements are made under the Safe Harbor provisions of the US Private Securities Litigation Reform Act of 1995.

在我們繼續之前，我想請各位參閱公司盈利新聞稿中的“安全港”聲明，該聲明也適用於本次電話會議，因為今天的電話會議將包含前瞻性聲明，包括 WeRide 的戰略和未來計劃。這些前瞻性陳述是根據 1995 年美國私人證券訴訟改革法案的安全港條款作出的。

Forward-looking statements involve inherent risks and uncertainties. The company's actual results could differ materially from those stated or implied by these forward-looking statements as a result of various important factors, and please refer to Risk Factors sections of the company's Form 20-F filed with the SEC and announcements on the website of the Hong Kong Stock Exchange for the full disclosure of these risk factors.

前瞻性陳述涉及固有的風險和不確定性。由於各種重要因素的影響，本公司的實際業績可能與這些前瞻性聲明中所述或暗示的業績存在重大差異，請參閱公司向美國證券交易委員會提交的 20-F 表格中的「風險因素」部分以及香港聯合交易所網站上的公告，以全面披露這些風險因素。

The company does not assume any obligations to update any forward-looking statements, except as required under applicable law. Please note that all numbers stated in management's prepared remarks are in RMB terms, and we will discuss non-IFRS measures today, which are more truly explained and reconciled to the most comparable measures reported in the company's earnings release and filings with the SEC and the Hong Kong Stock Exchange.

除適用法律要求外，本公司不承擔更新任何前瞻性聲明的義務。請注意，管理層準備的發言稿中所述的所有數字均以人民幣計價，我們今天將討論非國際財務報告準則（IFRS）指標，這些指標與公司盈利報告以及向美國證券交易委員會和香港聯合交易所提交的文件中所報告的最可比指標進行了更真實的解釋和核對。

With that, I'll now turn the call over to the company's Founder, Chairman and CEO, Dr. Tony Han. Please go ahead, sir.

接下來，我將把電話交給公司創辦人、董事長兼執行長東尼韓博士。請繼續，先生。

Thank you. Hello, everyone. Thank you for joining us today. I would like to begin by highlighting some of the key milestones we achieved this past quarter. Q3 was a period of extraordinary progress for WeRide. Most notably, we made history in Abu Dhabi by securing the world's first city-level, fully driverless robotaxi commercial permit outside the United States. And we will begin -- actually, we have already started the driverless operations through Uber, which I'm going to detail in the later slides this week.

謝謝。大家好。感謝您今天蒞臨。首先，我想重點介紹我們上個季度取得的一些關鍵里程碑。第三季是 WeRide 非凡進展的時期。最值得一提的是，我們在阿布達比創造了歷史，並獲得了美國以外全球首個城市級完全無人駕駛機器人計程車商業許可證。我們將開始——實際上，我們已經通過 Uber 開始了無人駕駛運營，我將在本週後面的幻燈片中詳細介紹。

With our recent expansion into Belgium and our inaugural driverless robotaxi license in Switzerland, WeRide has become the only company with autonomous driving permits for eight countries. By October, we have developed L4 fleets in 11 countries and more than 30 cities with over 1,600 L4 level autonomous driving vehicles in operation worldwide. Now let's take a look at our third quarter accomplishments. So let's turn into Abu Dhabi slides. Okay. Next slide, please.

隨著我們最近在比利時的業務拓展以及在瑞士獲得首個無人駕駛機器人計程車許可證，WeRide 已成為唯一一家在八個國家擁有自動駕駛許可證的公司。截至 10 月，我們已在 11 個國家和 30 多個城市建立了 L4 級自動駕駛車隊，全球共有 1,600 多輛 L4 級自動駕駛汽車投入營運。現在讓我們來看看第三季的成果。所以，讓我們化身為阿布達比滑梯吧。好的。請看下一張投影片。

So as mentioned earlier, WeRide has been officially approved to provide full driverless commercial robotaxi service in the UAE's capital, Abu Dhabi. This landmark authorization removes the requirements for in-car safety officer and demonstrates the regulators' strong confidence in our technology. Following this approval, WeRide and Uber collectively launched the region's first fully driverless fire charging robotaxi service this week, starting from Ya Island and with a citywide rollout underway.

如同前面所提到的，WeRide 已正式獲准在阿聯酋首都阿布達比提供完全無人駕駛的商業機器人計程車服務。這項具有里程碑意義的授權取消了對車內安全員的要求，並表明監管機構對我們技術的堅定信心。獲得批准後，WeRide 和 Uber 本週聯合推出了該地區首個完全無人駕駛的消防機器人計程車服務，該服務首先在亞島推出，目前正在全市範圍內推廣。

And next page. Our commercial operation at Abu Dhabi has begun in last December. Our service now covers roughly 50% of the city's core area. In half of 12 hours, our single vehicle can complete up to 20 trips per day. I think this is a quite exciting progress. In the third midterm -- in the midterm, we aim to extend our service hours to 24/7, increase vehicle utilization to more than 25 trips per day and improve human-to-vehicle ratio to 1:10. These numbers will lead us to a very healthy unit economics.

下一頁。我們在阿布達比的商業營運已於去年12月開始。我們的服務目前涵蓋了該市核心區域約 50% 的面積。在 12 小時的一半時間內，我們的一輛車每天最多可以完成 20 次行程。我認為這是一個令人振奮的進展。在第三個中期階段——在中期階段，我們的目標是將服務時間延長至全天候 24 小時，將車輛利用率提高到每天 25 次以上，並將人車比提高到 1:10。這些數據將使我們獲得非常健康的單位經濟效益。

We believe Abu Dhabi will set a global benchmark for large-scale and commercially viable robotaxi operation. And with all of these numbers, I think our unit economics is very, very healthy and can be profitable. And so I just want to emphasize this kind of breakthrough is quite exciting, and we work so hard for a whole year to achieve this full driverless robotaxi operation in Abu Dhabi. And this is the first city level outside of United States who are cable -- first city level robotaxi service out of the United States, and it is actually provided through Uber platform. So with all of these important factors, this is unparalleled, and we are so exciting that we are making the history.

我們相信阿布達比將為大規模、商業上可行的無人駕駛計程車營運樹立全球標竿。綜合所有這些數據，我認為我們的單位經濟效益非常非常健康，而且能夠獲利。因此，我只想強調，這種突破非常令人興奮，我們為此努力了一整年，才在阿布達比實現了完全無人駕駛機器人計程車的營運。這是美國以外第一個城市級別的有線機器人出租車服務——美國以外第一個城市級別的機器人出租車服務，實際上是透過 Uber 平台提供的。綜合所有這些重要因素，這是前所未有的，我們正在創造歷史，我們為此感到無比興奮。

Now let's talk about our current operation in Dubai. In September, we secured a self-driving vehicle trial permit from Dubai's Roads and Transport Authority and have begun road testing for our driverless operation in Dubai. Our goal is to launch supervised trial on Uber this year and the driverless commercial operation in the year of 2026, I mean, next year, we are going to provide driver robotax service in Dubai. Next page, please.

現在我們來談談我們在杜拜的當前營運情況。9 月，我們從杜拜道路和交通管理局獲得了自動駕駛車輛試驗許可證，並已開始在杜拜進行無人駕駛營運的道路測試。我們的目標是今年在 Uber 上啟動監督試驗，並在 2026 年實現無人駕駛商業營運。我的意思是，明年，我們將在杜拜提供無人駕駛計程車服務。請翻到下一頁。

Then we are going to talk our current operation in Saudi Arabia. In Riyadh, we began offering robotaxi rides through Uber in October, making our robotaxi service first and only publicly accessible robotax service in the Kingdom. With our development in the three largest cities in Middle East, that is Riyadh, Abu Dhabi and Dubai, we have more than 100 robotaxi vehicles in the Mid East region. The launch of driverless operation in Abu Dhabi is paving the way to scale the fleet to more than 500 vehicles by next year and to tens of thousands by 2030. So we are very excited and very confident and very proud and very confident and very optimistic about our full driverless robotaxi operation in Middle East.

接下來，我們將討論我們目前在沙烏地阿拉伯的業務運作。10 月，我們在利雅德開始透過 Uber 提供機器人計程車服務，使我們的機器人計程車服務成為沙烏地阿拉伯王國首個也是唯一一個面向公眾開放的機器人計程車服務。隨著我們在中東三大城市——利雅德、阿布達比和杜拜——的發展，我們在中東地區擁有超過 100 輛無人駕駛計程車。阿布達比無人駕駛營運的啟動，為明年將車隊規模擴大到 500 多輛，到 2030 年擴大到數萬輛鋪平了道路。因此，我們對在中東全面開展無人駕駛機器人計程車業務感到非常興奮、非常有信心、非常自豪、非常樂觀。

But that's not only our operation region, and I want to talk about East Asia and Europe. First, in Singapore, together with Grab, we received approval from the Land Transport Authority for both robotaxi and robobus in the area called the Pongo District. We plan to increase our EV test volume by 4 times by the end of this year. We also are integrating our technology into Grab's fleet management and routing system so that, in future, we can provide driverless robotaxi through Grab in Singapore, just what we have done in Abu Dhabi through Uber.

但這不僅限於我們的營運區域，我還想談談東亞和歐洲。首先，在新加坡，我們與 Grab 一起獲得了陸路交通管理局的批准，可以在名為 Pongo District 的地區運營機器人計程車和機器人巴士。我們計劃在今年年底前將電動車測試量增加 4 倍。我們還將我們的技術整合到 Grab 的車隊管理和路線規劃系統中，以便將來我們可以透過 Grab 在新加坡提供無人駕駛機器人計程車服務，就像我們在阿布達比透過 Uber 所做的那樣。

Actually, this -- all of these efforts will pay the groundwork for commercial service in the next phase. And then let's talk about Switzerland. In Europe, the expansion in Switzerland continues to lead our robotaxi deployment. We received the country's first driverless robotaxi license, enabling our autonomous operation in the fertile region. A full driverless public service is expected to be launched in the first half of 2026. That is our current operation in Europe and in East Asia and Singapore.

實際上，所有這些努力都將為下一階段的商業服務奠定基礎。然後我們來談談瑞士。在歐洲，瑞士的擴張繼續引領我們的無人駕駛計程車部署。我們獲得了全國首個無人駕駛機器人計程車牌照，使我們能夠在這一富饒的地區進行自主運作。預計2026年上半年將推出完全無人駕駛的大眾運輸服務。這就是我們目前在歐洲、東亞和新加坡的業務運作。

And now let's talk about China. In our China market, we continue to expand and innovate. As we scale our commercial fleets, we also launched a 24/7 driverless commercial service in Huangpu district at Guangzhou. This is an area of 150 square kilometers.

現在我們來談談中國。在中國市場，我們不斷拓展和創新。隨著我們擴大商用車隊規模，我們也在廣州黃浦區推出了全天候無人駕駛商用服務。這是一個面積為150平方公里的區域。

As of October, we have deployed more than 300 robotaxi in Guangzhou and over 100 in Beijing. For all of this service, you can hail a driverless robotaxi in this region through our WeRide Go app. User value is kept very close to our heart, and we recently introduced China's first free pickup and drop-off feature for robotaxi service we call (inaudible) service, allowing our system to intelligently recommend optimal boarding locations. This greatly improved both operational flexibility and user experience, which is well captured by our operational data.

截至10月，我們在廣州部署了300多輛無人駕駛計程車，在北京部署了100多輛。您可以透過我們的 WeRide Go 應用程式在該地區叫到無人駕駛機器人計程車，享受所有這些服務。用戶價值是我們非常重視的，我們最近推出了中國首個免費接送機器人出租車服務功能（我們稱之為（聽不清楚）服務），使我們的系統能夠聰明地推薦最佳上車地點。這大大提高了營運靈活性和用戶體驗，我們的營運數據也很好地反映了這一點。

In November, each robotaxi completed up to 25 daily trips in Guangzhou and 23 in Beijing, which is a clear evidence of accelerated adoption. And then this is our current exciting progress about robotaxi. Next, let's talk about our other applications.

11 月，每輛無人駕駛計程車在廣州每天完成多達 25 次行程，在北京每天完成多達 23 次行程，這清楚地表明了無人駕駛計程車技術的加速普及。接下來，我們將介紹我們在無人出租車領域目前取得的令人興奮的進展。接下來，我們來談談我們的其他應用。

First, robobus. Okay, next page. our robobus obtained Belgium's first Level 4 test permit, and we launched our operation in Luven, making Belgium the 11th country covered by our service. In Guangzhou, after serving more than 1 million passengers since 2021, we received an offer -- received an order for another -- for additional 100 midsized robobuses. This is a very exciting achievement. Actually, this is a newly developed robobus. In Hong Kong, we established a partnership with Guangzhong Bus Holding to deploy more than 500 Level 4 vehicles over the next three years.

首先是無人巴士。好的，下一頁。我們的無人駕駛巴士獲得了比利時首個 4 級測試許可，我們在盧文啟動了運營，使比利時成為我們服務覆蓋的第 11 個國家。在廣州，自 2021 年以來，我們已為超過 100 萬名乘客提供服務，之後我們又收到了一份訂單，要再增加 100 輛中型無人駕駛巴士。這是一項非常令人振奮的成就。實際上，這是一輛新開發的無人駕駛巴士。在香港，我們與廣中客車控股有限公司建立了合作關係，將在未來三年內部署 500 多輛 4 級車輛。

And for our L2+ level ADAS system, WeRide and Bosch achieved a major milestone in November with the start of production of WePilot 3.0. This is an end-to-end system. It's just like what Tesla has achieved through his -- through their FSD system. Our WePilot is totally comparable to what Tesla can do with FSD. The WePilot 3.0 will debut with a refreshed Cherry Exceed ES and ET model and existing owners will receive OTA upgrades. With this kind of new feature, every owner of Cherry Exceed ES and ET can enjoy the experience of Tesla's FSD.

對於我們的 L2+ 級 ADAS 系統，WeRide 和 Bosch 在 11 月實現了一個重要的里程碑，WePilot 3.0 開始量產。這是一個端到端的系統。這就像特斯拉透過他們的FSD系統所取得的成就。我們的 WePilot 功能與特斯拉的 FSD 功能完全可以媲美。WePilot 3.0 將率先推出新款 Cherry Exceed ES 和 ET 車型，現有用戶將獲得 OTA 升級。有了這項新功能，每位 Cherry Exceed ES 和 ET 的車主都可以享受特斯拉 FSD 的體驗。

With this exceptional end-to-end system, WePilot has also been selected as the major ADAS system provider by Guangzhou Automotive Group, GAC, for several of their passenger car models. So that part is actually a very exciting progress, demonstrating WeRide can -- are not only capable of doing L4 level robotaxi, but also are capable of doing L2++ level ADAS for massive production car.

憑藉這款卓越的端到端系統，WePilot 也被廣州汽車集團（廣汽集團）選為旗下多款乘用車的主要 ADAS 系統供應商。所以這部分實際上是一個非常令人興奮的進展，它證明了 WeRide 不僅能夠實現 L4 級自動駕駛出租車，而且還能夠為大規模生產的汽車實現 L2++ 級高級駕駛輔助系統 (ADAS)。

And this page actually shows -- summarize our footprint in the global. Actually, WeRide's strategy prioritize a balanced development in the global market. I want to explain this slide a little bit. You can see like in these 11 countries, we have different levels of operation. We have tested or we can operate it without the driver. And so they are showing in the legend.

本頁實際上展示了——概括我們在全球的影響力。事實上，WeRide的策略優先考慮在全球市場實現均衡發展。我想稍微解釋一下這張投影片。你可以看到，在這11個國家，我們的營運水準各不相同。我們已經測試過，即使沒有司機也能操作它。所以它們都顯示在圖例中。

So you see our multiproduct offering has maximized the value of our strategy, making us the only company whose technology is available in the 11 countries shown here. So we actually have a wide spectrum of applications and service available for the global market.

因此，您可以看到，我們的多產品組合最大限度地發揮了我們策略的價值，使我們成為唯一一家技術在圖中所示的 11 個國家/地區均可使用的公司。因此，我們實際上為全球市場提供廣泛的應用和服務。

Next, let me discuss about the backbone of our technology. This is called WeRun 1 Universal platform. By starting from supporting L4 applications alone in the early days, WeRun 1 has grown gradually grown into a more powerful platform that empowers the full spectrum from L2 to L4 while continuously breeding new tools and systems. So one of the most preeminent is our world model, we write Genesis.

接下來，我想談談我們技術的核心。這被稱為WeRun 1通用平台。WeRun 1 最初只支援 L4 應用，後來逐漸發展成為一個更強大的平台，能夠支援從 L2 到 L4 的整個應用範圍，同時不斷湧現新的工具和系統。因此，其中最重要的一個就是我們的世界模型，我們寫了《創世記》。

So in our Genesis model, you can see Genesis is a new platform and will allow autonomous vehicles to be tested in a digital twin of the real world safely, efficiently and at a large scale. It features the data loop, algorithm loop and simulation validation loop that are essential for scalable autonomy. This is our world model, and it's seamlessly integrated with our end-to-end system.

因此，在我們的 Genesis 模型中，您可以看到 Genesis 是一個新平台，它將允許在真實世界的數位孿生體中安全、高效、大規模地測試自動駕駛汽車。它具有資料循環、演算法循環和模擬驗證循環，這些對於可擴展的自主性至關重要。這是我們的世界模型，它與我們的端到端系統無縫整合。

So this is a unique technology advantage. Our world model can be seamlessly integrated into end-to-end ADAS system, and our L4 system can leverage on the data we collected from our L2 level data. So this Genesis form as the core flywheel, actually, we call it a double flywheel. We can actually leverage on L2++ level massive production car data to improve our robotaxi. In turn, our robotaxi data with redundancy can help us to boost the performance of our ADAS system.

所以這是一項獨特的技術優勢。我們的世界模型可以無縫整合到端到端ADAS系統中，我們的L4系統可以利用我們從L2級資料收集的資料。所以，這個以 Genesis 為核心的飛輪，實際上，我們稱之為雙飛輪。我們實際上可以利用 L2++ 等級的海量量產車數據來改善我們的無人計程車。反過來，我們具有冗餘性的機器人計程車數據可以幫助我們提高 ADAS 系統的效能。

I want to emphasize in this world, there's only one company, WeRide, can do so. On one hand side, we have a large-scale robotaxi fleet make WeRide capable of doing -- leading robotaxi service like Waymo and other companies like they can do in the US. So that we can actually capture all the characteristics of full driverless operation. At the same time, WeRide supply WePoint 3.0 and very advanced ADAS system comparable to the FSD of Tesla, and we can leverage on the mass production car data and collect all this kind of data in a very broad sense in all kinds of scenario to help us to improve the performance of robotaxi.

我想強調的是，在這個世界上，只有一家公司，WeRide，能夠做到這一點。一方面，我們擁有大規模的無人駕駛計程車車隊，這使得 WeRide 能夠像 Waymo 和其他公司在美國一樣，提供領先的無人駕駛計程車服務。這樣我們才能真正捕捉到完全無人駕駛操作的所有特徵。同時，WeRide 提供 WePoint 3.0 和非常先進的 ADAS 系統，可與特斯拉的 FSD 相媲美，我們可以利用量產車數據，在各種場景下廣泛收集此類數據，以幫助我們提高無人駕駛計程車的效能。

We believe we can combine the benefit of L4 and L2+. This hybrid architecture enhance adaptability, reliability, safety and transparency, ultimately enabling robust commercial deployment. We look forward to sharing more about this advantage soon. In summary, Q3 was a quarter of exceptional execution. We expand our global leadership, and we translated technology innovation into commercial reality.

我們相信我們可以結合 L4 和 L2+ 的優勢。這種混合架構增強了適應性、可靠性、安全性和透明度，最終實現了穩健的商業部署。我們期待盡快與大家分享更多關於這項優勢的資訊。總而言之，第三季執行情況非常出色。我們擴大了全球領先地位，並將技術創新轉化為商業現實。

With that, I will hand over the call over to our CFO, Jennifer, to discuss our financial performance. Jennifer, please go ahead to discuss about the financial numbers.

接下來，我將把電話交給我們的財務長詹妮弗，由她來討論我們的財務表現。珍妮弗，請繼續討論財務數據。

Thank you, Tony. Hello, everyone. Before we dive into the third quarter financials, I want to highlight that all figures are in RMB and comparisons are year-over-year unless otherwise stated. Now let's discuss our third quarter financial performance.

謝謝你，托尼。大家好。在深入探討第三季財務數據之前，我想強調的是，除非另有說明，所有數字均以人民幣計價，比較均為同比。現在我們來討論一下第三季的財務表現。

We delivered total revenue of RMB171 million with a year-over-year growth of 144%, driven by our continued fleet expansion and increase in service penetration. The revenue growth also reflects a significant milestone we have achieved during this quarter, supported by our advanced technology, robust deployment and operational capabilities.

在持續擴大車隊規模和服務滲透率提高的推動下，我們實現了總收入1.71億元人民幣，年增144%。營收成長也反映了我們在本季取得的一個重要里程碑，這得益於我們先進的技術、強大的部署和營運能力。

Our revenue came from both product revenue and service revenue. Product revenue delivered strong growth of 428% to RMB79 million in this quarter, an encouraging result driven by the increased sale of our robotaxi and robobuses. Service revenue grew 67% to RMB92 million in Q3, supported by an increase of RMB29 million from intelligent data service and an increase of RMB8 million in autonomous driving-related operational and technical services. Service revenue has surpassed product revenue in this quarter, demonstrating a continued growth momentum and healthy business structure.

我們的收入來源包括產品收入和服務收入。本季產品營收強勁成長 428% 至 7,900 萬元，這項令人鼓舞的業績主要得益於我們的無人計程車和無人巴士銷售的成長。第三季服務收入成長67%至人民幣9,200萬元，其中智慧數據服務收入成長人民幣2,900萬元，自動駕駛相關營運及技術服務收入成長人民幣800萬元。本季服務收入超過產品收入，顯示公司持續保持成長勢頭，業務結構健康。

Among our product lines, what really stood out in Q3, same as in the last two quarters, was our robotaxi businesses. Robotaxi revenue increased 761% year-over-year to RMB35 million in Q3, accounting for 21% of total revenue in this quarter. With our new federal permits in UAE, we are the first and only robotaxi company that have begun full driverless robotaxi operation in UAE. Removing in-car safety officer is a critical milestone from a financial perspective, which will enable our robotaxi service to achieve unit economic breakeven.

在我們的產品線中，與前兩季一樣，第三季表現最突出的還是我們的無人駕駛計程車業務。第三季度，Robotaxi 營收年增 761% 至 3,500 萬元，佔本季總營收的 21%。憑藉我們在阿聯酋獲得的新聯邦許可證，我們是阿聯酋第一家也是唯一一家開始全面開展無人駕駛機器人計程車營運的公司。從財務角度來看，取消車內安全員是一個重要的里程碑，這將使我們的無人駕駛計程車服務實現單位經濟盈虧平衡。

The quality of our growth is also compelling. Group level gross profit increased 1,124% to RMB56 million for the third quarter, with a group level gross margin of 33%, demonstrating our industry-leading gross margin as our business continue to grow. We aim to keep delivering business value along with our globalization strategy. Operating expense decreased 51% to RMB436 million, with R&D expense accounting for 73% of the total operating expenses.

我們成長的品質也令人矚目。集團第三季​​毛利成長1124%至人民幣5,600萬元，集團毛利率為33%，隨著業務的持續成長，集團毛利率處於業界領先地位。我們的目標是在推動全球化策略的同時，持續創造商業價值。營業費用下降51%至人民幣4.36億元，其中研發費用佔營業總費用的73%。

To break down further, R&D expense increased by 24% to RMB316 million in the third quarter of 2025 compared to the same period of 2024. Excluding share-based compensation, R&D expense grew 39% to RMB288 million as we further strengthened our global data compliance and advanced R&D efforts for our pre-installed robotaxi.

具體來說，2025 年第三季研發支出比 2024 年同期成長 24%，達到 3.16 億元。不計股權激勵，研發費用成長 39% 至 2.88 億元，因為我們進一步加強了全球數據合規性，並推進了預載無人計程車的研發工作。

The increase in R&D expense was primarily due to an increase of RMB31 million in service fee for R&D projects, an increase of RMB21 million in personnel-related expense from headcount increase and an increase of RMB23 million in material consumption and depreciation and amortization expenses.

研發費用增加主要是因為研發項目服務費增加人民幣3100萬元，人員增加導致人員相關費用增加人民幣2100萬元，以及材料消耗和折舊攤銷費用增加人民幣2300萬元。

Administrative expense decreased by 84% to RMB100 million in the third quarter of 2025 compared to the same period in 2024. Excluding share-based compensation, administrative expense increased by 23% to RMB74 million. The increase was primarily due to an increase of RMB6 million in professional service fee, mainly related to legal compliance service and an increase of RMB4 million in personnel costs as we continue to build necessary supporting functions to grow our business.

2025年第三季行政費用較2024年同期減少84%，至人民幣1億元。不計股權激勵費用，行政費用增加23%至人民幣7,400萬元。此次成長主要是由於專業服務費增加了人民幣600萬元，主要與法律合規服務有關；以及人員成本增加了人民幣400萬元，因為我們繼續建立必要的支援職能以發展我們的業務。

Selling expenses increased 23% to RMB19 million in the third quarter of 2025 compared to the same period of 2024. Excluding share-based compensation, selling expense increased by 36% to RMB19 million, which was well below the sales increase. Our commitment to R&D is the backbone of our strategy. We will continue to direct our resource there to pioneer the industry innovation and keep building our competitive advantage. Alongside this, we will strategically grow our global team with a clear focus on region that has accelerated the adoption of L4 solutions. This ensures that we have a world-class talent needed to support our business expansion.

2025年第三季銷售費用較2024年同期成長23%，達1,900萬元人民幣。不計股權激勵費用，銷售費用成長36%至人民幣1,900萬元，遠低於銷售額的成長。我們對研發的投入是我們策略的基石。我們將繼續投入資源到那裡，引領產業創新，並不斷增強我們的競爭優勢。同時，我們將策略性地發展我們的全球團隊，並專注於加速採用 L4 解決方案的地區。這確保我們擁有世界一流的人才，以支持我們的業務擴張。

Our net loss narrowed by 71% to [RMB307 million] in the third quarter of 2025. On a non-IFRS basis, adjusted net loss increased 15% to RMB276 million, largely due to an ongoing R&D investment and broader operational support required for the expansion of our business.

2025年第三季度，我們的淨虧損收窄71%，至人民幣3.07億元。以非國際財務報告準則計算，經調整後的淨虧損增加 15% 至人民幣 2.76 億元，主要原因是持續的研發投資以及為業務擴張所需的更廣泛的營運支援。

As of September 30, 2025, we had RMB4.5 billion in cash and cash equivalents and time deposits, RMB926 million in investment in wealth management products and RMB18 million in restricted cash. We had short-term bank borrowing of RMB245 million. Our current liquidity reserve, along with the proceeds from our recent Hong Kong due primary listing in November have enabled us with a resilient position for our R&D-focused strategy and our globalization deployment progress.

截至2025年9月30日，我們持有現金及現金等價物及定期存款人民幣45億元，財富管理產品投資人民幣9.26億元，受限現金人民幣1800萬元。我們有2.45億元的短期銀行貸款。我們目前的流動資金儲備，加上我們最近於 11 月在香港首次公開募股獲得的收益，使我們擁有了強大的實力，能夠以研發為重點的戰略和全球化部署進程為支撐。

Our fully driverless robotaxi commercial permit in Abu Dhabi is not just a local milestone. It's a scalable blueprint for the global industry. It demonstrates a viable path for city level full driverless operation outside the US, along with the potential for profitable unit economics in major international markets. Our strategy is to scale this model globally. We have the complete package, the technology, the operational experience, a proven safety record and regulatory trust. In the next five years, we will achieve large-scale L4 deployment, creating a sustainable business and delivering tremendous value of autonomous driving to the shareholders.

我們在阿布達比獲得的完全無人駕駛機器人計程車商業許可證不僅僅是一個地方性的里程碑。這是一個適用於全球產業的可擴展藍圖。它展示了在美國以外城市層面實現完全無人駕駛營運的可行路徑，以及在主要國際市場實現盈利性單位經濟效益的潛力。我們的策略是將這種模式推廣到全球。我們擁有完整的資源，包括技術、營運經驗、可靠的安全記錄和監管機構的信任。未來五年，我們將實現L4級自動駕駛的大規模部署，打造永續發展的業務，並為股東創造巨大的自動駕駛價值。

With that, operator, we are now ready to take on some questions.

操作員，那我們現在可以回答一些問題了。

(Operator Instructions)

（操作說明）

Tim Hsiao, Morgan Stanley.

Tim Hsiao，摩根士丹利。

Hi, Tony and Jennifer, thanks for taking my question. This is Tim from Morgan Stanley. Congratulations on the strong results and continuous extension and robotaxi operations globally. I have two questions. The first question, we noticed that we have officially started commercial deployment of driverless robotaxi in Abu Dhabi, UAE. So in addition to the volume upsize to revised fleet sales, as Tony just mentioned, How should we quantify the revenue opportunities of a vehicle sales, revenue charge, and the profit sharing in the long run? That's my first question.

嗨，Tony 和 Jennifer，謝謝你們回答我的問題。我是摩根士丹利的提姆。恭喜你們取得優異成績，並在全球範圍內持續拓展和經營機器人計程車業務。我有兩個問題。第一個問題是，我們注意到我們已經在阿聯酋阿布達比正式開始商業部署無人駕駛機器人計程車。所以，除了像托尼剛才提到的那樣，透過擴大規模來調整車隊銷售之外，我們應該如何量化車輛銷售的收入機會、收入費用以及長期的利潤分成呢？這是我的第一個問題。

Okay. Thank you, Tim. That's a great question. I'll take the first one. So for the benefit of all listeners, I'd like to briefly elaborate on our robotaxi's business model. In domestic China, we mainly own and operate vehicles by ourselves and on our own ride-hailing platform, WeRide Go. So before -- after the UE gets to breakeven point in the next few years in China, we will gradually engage third-party asset owners and partnership to them. And for now, we pretty much own like all the vehicles by ourselves.

好的。謝謝你，提姆。這是一個很好的問題。我選第一個。為了所有聽眾的利益，我想簡要地詳細介紹一下我們無人出租車的商業模式。在中國國內，我們主要擁有並經營自有車輛，並透過我們自己的叫車平台WeRide Go提供服務。因此，在未來幾年內，當UE在中國實現損益平衡之後，我們將逐步引入第三方資產所有者並與其建立合作關係。目前，我們幾乎所有的車輛都歸我們自己所有。

And international market is different. From day one, we collaborate with platform partners such like Uber, Grab, SVB, TXAI, and we generate revenue from three main streams, three streams. The first one is revenue share from the ride fare and second one is the annual licensing and third is the sale of the vehicle. So vehicle sale is considered as the product revenue. WeRide can scale up the robotaxi fleet much quicker and in a lighter business model like on an asset basis since the robotaxi operation fleet doesn't sit on our own balance sheet.

國際市場則有所不同。從一開始，我們就與 Uber、Grab、SVB、TXAI 等平台合作夥伴合作，並透過三大主要管道創造收入。第一部分是車費收入分成，第二部分是年度牌照費，第三部分是車輛銷售收入。因此，車輛銷售額被視為產品收入。WeRide 可以更快地擴大機器人計程車車隊規模，並採用更輕量級的商業模式（例如基於資產的商業模式），因為機器人計程車營運車隊不屬於我們自己的資產負債表。

We really just sell this to our partner already. And the revenue share and annual licensing of the recurring service revenue over the whole lifespan of the vehicle, which tend to be five to seven years. In particular, revenue share will become a significant multiplier following the expansion of the fleet size. We'll take Middle as an example. a robotaxi at a human level utilization, which means they can complete like 25 orders per day can generate an annual like revenue of over USD90,000 like on the platform.

我們其實已經把這個賣給我們的合作夥伴了。以及車輛整個生命週期（通常為五到七年）內經常性服務收入的收入分成和年度許可費。特別是，隨著船隊規模的擴大，收入份額將成為一個重要的倍增器。我們以 Middle 為例。一輛機器人計程車如果能達到人類計程車的使用率水平，也就是每天可以完成大約 25 個訂單，那麼它在平台上每年就能產生超過 9 萬美元的收入。

If WeRide take 30% of the revenue share, that will give us USD30,000 per car per year as service -- as revenue share. If we can take 70% of the revenue share, that will give us like USD60,000 per car per year. So if we are moving this one step closer to the goal and to see what we have already, let's say, in Abu Dhabi, right now, we have a significant presence with near 100 robotaxi in Abu Dhabi. Now we already cover 50% of the city core area. And the commercial model is we're integrating on platform like Uber. Right now, we are charging at the same price level at Uber X and Uber Comfort.

如果 WeRide 抽取 30% 的收入分成，那麼我們每年每輛車將獲得 30,000 美元的服務費——作為收入分成。如果我們能獲得 70% 的收入分成，那我們每輛車每年就能賺到大約 6 萬美元。所以，如果我們朝著目標又邁進了一步，看看我們已經取得了什麼成就，比如說，在阿布達比，目前我們已經擁有了近 100 輛無人駕駛計程車，在阿布達比擁有相當大的市場份額。現在我們已經涵蓋了市中心區域的50%。我們的商業模式是整合到像 Uber 這樣的平台。目前，Uber X 和 Uber Comfort 的收費標準相同。

In fact, if you get like get on the Uber and to call the robot -- just to call normal ride-hailing car in half of the city, no matter you pick Uber X, Uber Comfort or autonomous option, you can all get a vehicle. And this demonstrates that our service is competitive with the mainstream like ride-hailing from day one on the pricing level.

事實上，如果你像在城市裡那樣，乘坐 Uber 並呼叫機器人——或者只是在半個城市裡呼叫普通的叫車，無論你選擇 Uber X、Uber Comfort 還是自動駕駛選項，你都能得到一輛車。這表明，從一開始，我們的服務在價格方面就與主流的叫車服務具有競爭力。

And unit economic is -- on the unit economic side, the most critical metric is utilization. Right now, we already achieve a daily average like 12 order per vehicle in a 12-hour shift. Sometimes we can get to -- on the good days, we can get to more than 20 orders per vehicle per day for the 12-hour shift. And it's already indicating a strong user preference and stickiness. So for the information, for 12 orders per vehicle per day, we can already get to the breakeven threshold in this market.

而單位經濟效益方面，最關鍵的指標就是利用率。目前，我們每輛車在 12 小時輪班期間，平均每天已經能完成 12 個訂單。有時，在順利的日子裡，我們每輛車每天在 12 小時的輪班時間內可以完成 20 多個訂單。這已經表明用戶對其有很強的偏好和黏性。因此，就目前的資訊來看，如果每輛車每天能接到 12 個訂單，我們在這個市場就已經可以達到損益平衡點了。

So there's huge profitability potential. Based on our current driverless cost structure and plan to extend the hours to 24 hours next year, we project an average daily order can reach to 25 per vehicle per day. And this level of utilization will lead to a very strong profitability potential next year. So with the -- while the specific percentage of revenue share is confidential between different partners and -- but this approach can empower a sustainable win-win partnership for everyone in the ecosystem.

因此，其中蘊藏著巨大的獲利潛力。根據我們目前的無人駕駛成本結構以及明年將營運時間延長至 24 小時的計劃，我們預計平均每天每輛車的訂單量可達 25 件。而這樣的利用率將為明年帶來非常強勁的獲利潛力。因此，雖然特定的收入分成比例是不同合作夥伴之間的機密信息，但這種方法可以為生態系統中的每個人建立可持續的雙贏夥伴關係。

My second question is also related to WeRide global business. So looking forward, in addition to operations in Abu Dhabi and Switzerland, which we just announced, which market could step up as a key volume driver to WeRide? And do we really need to accelerate R&D and selling expanding more aggressively into next year 2026 to finance the company's robust expansion in overseas? That's my second question.

我的第二個問題也與 WeRide 的全球業務有關。展望未來，除了我們剛剛宣布的在阿布達比和瑞士的業務之外，哪個市場有可能成為 WeRide 的關鍵銷售驅動因素？我們真的需要加快研發和銷售擴張步伐，在2026年更積極地拓展業務，為公司在海外的強勁擴張提供資金嗎？這是我的第二個問題。

Okay. I'll take the question. And so first of all, I think, so besides Abu Dhabi, we plan which markets would step up as key volume driver. So to us, in our plan like so first of all, in the Mid East, we have two major cities in UAE, Dubai and Abu Dhabi currently is already. We have already got the permits, and we are doing extensive road testing.

好的。我來回答這個問題。所以首先，我認為，除了阿布達比之外，我們還要規劃哪些市場將成為關鍵的銷售驅動因素。所以對我們來說，在我們的計畫中，首先，在中東，我們在阿聯酋有兩個主要城市，杜拜和阿布達比，目前已經是這樣了。我們已經拿到了許可證，正在進行廣泛的道路測試。

And UAE that make one country, we pay a lot of attention to. It's a very important market. And there's also Saudi Arabia. So you can see major countries in South and Mid East. So they are part of potential countries. And also Europe, and also -- and other developed countries in Asia like Japan, Singapore and Korea, they are all potential markets can help us to drive the volume up.

而阿聯酋作為一個國家，我們非常關注它。這是一個非常重要的市場。還有沙烏地阿拉伯。所以你可以看到南亞和中東的主要國家。所以它們是潛在國家的一部分。此外，歐洲以及亞洲其他已開發國家，如日本、新加坡和韓國，都是潛在的市場，可以幫助我們提高銷售量。

But one thing I want to point out is like Tim, you just asked a very good question that it's also I have been thinking about this over the years all the time that is what are our target markets, and which markets can make -- can we make our service and products very profitable? And I think through our tested or through our operation in Abu Dhabi, we find something so-called Abu Dhabi model, okay? Together with Uber, we found the unique economics is good and give us a very promising projection that we will soon -- in this region, we can make a good, very profitable service. This Abu Dhabi model actually created a road map for other cities. So with Uber, and I think we will try to copy this kind of model to the similar cities.

但我想指出的一點是，就像 Tim 剛才問的那個非常好的問題一樣，這也是我多年來一直在思考的問題，那就是我們的目標市場是什麼，哪些市場能夠——我們能否讓我們的服務和產品非常有利可圖？我認為，透過我們在阿布達比的實踐或運營，我們找到了所謂的阿布達比模式，好嗎？我們與 Uber 合作發現，這種獨特的經濟模式非常有利可圖，並為我們帶來了非常有前景的預測，我們很快就能在這個地區提供優質且非常有利可圖的服務。阿布達比模式實際上為其他城市提供了借鏡。所以，有了 Uber，我認為我們會嘗試將這種模式複製到類似的城市。

And also this is a combination of our current technology, our strategy as compared with our -- and also our collaboration with our strategy, strategic partnership. So with this model, I think we tend to copy to Singapore by by former another alliance with [Grab] also a very important strategic partner to do it in Singapore and also potentially on East Asia. And in China, we are focusing on developing a robo taxi service based on our own applications. And about expansion although we are increasing R&D investment to build stronger technology platforms, and we also quite hard to recruit top tenants, but we expect our growth of related expenditure to be moderate because we want to adopt a satellite model to strike a balance between scaling and the investment.

此外，這也是我們當前技術、我們的策略與我們的——以及我們與我們的策略、策略夥伴關係的合作的結合。所以，我認為我們傾向於透過與[Grab]的另一個聯盟來複製這種模式，Grab也是一個非常重要的策略合作夥伴，以便在新加坡以及潛在的東亞地區開展這項業務。在中國，我們正專注於開發基於我們自主應用程式的無人駕駛計程車服務。關於擴張，雖然我們正在增加研發投入以建立更強大的技術平台，而且我們也很難招募頂尖租戶，但我們預計相關支出的成長將保持適度，因為我們希望採用衛星模式，在規模和投資之間取得平衡。

One of our very midterm goal is trying to reach the profitability at the same time, maintain our strong market share. And also, we still want to innovate. So we have to strike a balance between investment and expenditure and the development. But with our current progress, I think I'm very optimistic because I have already seen the progress of the Abu Dhabi model. And what's next to do is trying to find all the places we can easily copy our Abu Dhabi model to, and by gathering all strengths from these potential markets, we want to achieve profitability in the near future.

中期目標之一是在實現盈利的同時，保持我們強大的市場份額。而且，我們仍然希望進行創新。因此，我們必須在投資、支出和發展之間取得平衡。但就我們目前的進展而言，我非常樂觀，因為我已經看到了阿布達比模式的成效。接下來我們要做的就是努力尋找所有可以輕鬆複製阿布達比模式的地方，並透過匯集這些潛在市場的所有優勢，我們希望在不久的將來實現盈利。

Alex Yao, JP Morgan.

Alex Yao，摩根大通。

I have two questions. Number one, what is your take of the robotaxi business in China? How do you envision economics to change in the future for China and for international market, respectively? The second question is how quickly can the driveless milestone of the Abu Dhabi operation be replicated in other markets? What can we expect for your fleet expansion plan globally? And what are the catalysts or hurdles for your plan?

我有兩個問題。首先，您如何看待中國的無人駕駛計程車產業？您認為未來中國經濟和國際市場的經濟將分別發生什麼樣的變化？第二個問題是，阿布達比無人駕駛營運的里程碑能否迅速在其他市場複製？貴公司在全球的機隊擴張計畫有哪些預期？那麼，你的計劃有哪些催化劑或障礙？

Okay. Let me try to answer these questions one by one. Although it's claimed to be two questions, it's two group of questions. So first question, if I remember clearly, it's roughly about our -- what do we think about China market and what's our plan for China market and our thoughts on the economics of China market? So first of all, I have spoken a lot about the global market, Mid East, Asian market, East Asian market, European market, and but one of our China, okay? China, our headquarter is located in China. China is stepping in a major market. And we also -- definitely the China is one of the most important markets we are targeting at, okay?

好的。讓我來逐一回答這些問題。雖然表面上是兩個問題，但實際上是兩組問題。第一個問題，如果我沒記錯的話，大致上是關於我們對中國市場的看法，我們對中國市場的計畫是什麼，以及我們對中國市場經濟的看法？首先，我談了很多關於全球市場、中東市場、亞洲市場、東亞市場、歐洲市場，但我們中國市場，好嗎？我們的總部位於中國。中國正在進軍一個重要市場。而且，中國絕對是我們重點瞄準的最重要的市場之一，好嗎？

Now let me elaborate about the pros and cons to put resource in China market and what our strategy is. So first of all, China is a unique market with the largest user base and the dynamic economics. And also, it is a great test ground. It has been both our technology proving ground and ideal operation, our sandbox for our very innovative ideas. But of course, that doesn't mean like we treat them as a lab, okay?

現在讓我詳細闡述一下在中國市場投入資源的利弊以及我們的策略。首先，中國是一個獨特的市場，擁有最大的使用者群體和充滿活力的經濟。而且，它也是一個很好的試驗場。它既是我們的技術試驗場，也是理想的營運場所，也是我們實踐創新理念的沙盒。當然，這並不代表我們要把它們當成實驗室來對待，好嗎？

We want to -- while we do a robotaxi operation or trial operation, we keep safety as our top priority. But still, with all kinds of scenario, all kinds of different climate, different weather conditions, china is a vast country, and we actually -- we tested for an different methods, different algorithms. So that part actually China is unique and a very big market.

我們希望－在進行無人駕駛計程車營運或試營運時，始終將安全放在首位。但是，考慮到各種不同的情況、不同的氣候、不同的天氣條件，中國是一個幅員遼闊的國家，我們實際上——我們測試了不同的方法、不同的演算法。所以，中國在這方面確實很獨特，而且是一個非常大的市場。

But we believe -- but China is also very -- in terms of development, it's not that balanced. They have Tier 1 city, looks like Paris and New York. There's a Tier 4 city likes like rural area. So we believe profitability in Tier 1 cities can be achieved with a combination of three elements. Number one, city level drivers permit. Number two, average daily order of high double digits.

但我們認為——但中國在發展方面也非常——並不那麼平衡。他們擁有像巴黎和紐約那樣的一線城市。有一個四線城市，但感覺卻像農村地區。因此，我們認為，一線城市的獲利能力可以透過三個要素的結合來實現。第一，市級駕駛許可證。第二，平均每日訂單量達到兩位數以上。

That means like what we try to achieve in Middle East should be achieved in China also, more than 20 orders per day. Number three, kind of a relatively healthy price, okay? Although these days, the taxi fares in China is still relatively low, but we expect to see the fare to grow a little bit. So it's still kind of relatively healthy. So with these three factors, we believe we still need to expand our market in China, mainly in Tier 1 cities.

這意味著我們在中東努力實現的目標也應該在中國實現，即每天超過 20 個訂單。第三，價格還算合理，好嗎？雖然目前中國的計程車費用仍然相對較低，但我們預計費用會略有上漲。所以它仍然算是比較健康的。綜上所述，我們認為仍需在中國拓展市場，尤其是在一線城市。

So far, we have achieved like, as I mentioned before, right, 300 robotaxi in Guangzhou and another 100 -- more than 100 robotaxi in Beijing in the areas of 150 kilometers area, and we are continuing to expand that. And although we have implemented a free product and drop off features help us to improve the user experience. And I think we aim at supply better robotaxi service than traditional taxi service supplied by human driver. Therefore, we can get more orders, and also we can give better user experience so that we can be ordered more frequently.

正如我之前提到的，到目前為止，我們在廣州已經投放了 300 輛無人駕駛計程車，在北京方圓 150 公里的範圍內投放了 100 多輛無人駕駛計程車，而且我們還在繼續擴大投放範圍。儘管我們已經實施了免費產品和退貨功能，但這有助於我們改善用戶體驗。我認為我們的目標是提供比傳統人類司機駕駛的計程車服務更好的機器人計程車服務。因此，我們可以獲得更多訂單，也可以提供更好的使用者體驗，從而提高訂單量。

And we expect that the economics of all of these markets will help to improve over time. And also, we want to actually learn what we have in China to expand to -- and learn what we have learned in China and yield them as our competitive advantage in the global market. Therefore, I think we still treat China as one of the most important markets, and we will keep on invest and inject resource in this market.

我們預計，隨著時間的推移，所有這些市場的經濟狀況都會有所改善。此外，我們也想了解我們在中國有哪些可以拓展的業務——並學習我們在中國學到的東西，並將它們轉化為我們在全球市場上的競爭優勢。因此，我認為我們仍然將中國視為最重要的市場之一，我們將繼續在這個市場進行投資和注入資源。

Okay. And the second good question is about how quickly we can copy Abu Dhabi model to the rest of the world, okay? First of all, thanks for asking this question, Andy. And the -- we believe we have find this kind of Abu Dhabi model. It's kind of like a native effect.

好的。第二個好問題是，我們能以多快的速度將阿布達比模式複製到世界其他地方，好嗎？首先，感謝你提出這個問題，安迪。而且──我們相信我們已經找到了這種阿布達比模式。這有點像是一種原生效果。

And WeRide has kind of a unique first mover because we just got a city level drivers permit. It is the only one so far out of the United States, that you can have a city drives permit and you can provide the service through Uber. So that means make the service volume equity available. Therefore, I think we can quickly copy to some similar market like Dubai in UAE same country. We are in Saudi and same region, and Singapore with a grab support.

WeRide 算是比較獨特的先行者，因為我們剛剛獲得了市級駕駛許可證。這是目前美國以外唯一一個可以讓你獲得城市駕駛許可證並透過 Uber 提供服務的城市。所以這意味著要讓服務量權益得以實現。因此，我認為我們可以迅速將這種模式複製到阿聯酋等類似國家的市場。我們位於沙烏地阿拉伯和新加坡，均在同一地區，並有 Grab 提供支援。

So we are trying to come to these kind of countries. And I believe the regulatory condition, all of other factors are kind of similar. And we also want to emphasize Europe is a very important market, and we are trying to see when we can comp it to Europe.

所以我們正努力前往這類國家。我認為監管條件和其他所有因素都大同小異。我們也要強調，歐洲是一個非常重要的市場，我們正在努力尋找與歐洲市場比較的方法。

And about the catalysts and hurdles, capless actually is because of this massive effect or other countries are more prone to to allow our operations are more prudent to give us permit. But at the same time, the hurdle is still the regulatory issues. We want to make sure we use our -- we leverage our current successful experience to get more driverless permits so that we can deploy the service. That's all I want to say about this question.

至於催化劑和障礙，無上限實際上是由於這種巨大的影響，或者其他國家更傾向於允許我們的運營，更謹慎地給我們頒發許可證。但同時，監管問題仍是最大的障礙。我們希望確保利用我們目前成功的經驗來獲得更多無人駕駛許可證，以便我們能夠部署這項服務。關於這個問題，我就想說這些。

Ming-Hsun Lee, Bank of America.

李明勳，美國銀行。

I have two questions as well. So first question, we are seeing more OEMs and ride-hailing companies announcing to plan enter on the robotaxi business. What are the WeRide's key advantages? And how should we think about the competitive landscape in the future? That's my first question.

我也有兩個問題。第一個問題是，我們看到越來越多的汽車製造商和叫車公司宣布計劃進軍無人駕駛計程車產業。WeRide的主要優勢是什麼？那麼，我們該如何看待未來的競爭格局呢？這是我的第一個問題。

Okay. So I think these days because of the increasing discussion an increasing maturity of robotaxi, you see so many car OEMs and platforms -- car-hailing platform start to talk about robotaxi start announce their local tax fee strategy. But one thing I want to mention is like robotaxi -- to do robotaxi is not easy, okay? It takes many years of efforts, technology accumulation, regulatory exploration, and that's why there are so few mature robotaxi company in this world. If you come to mature robotaxi service, one then they have open to public driverless robotaxi operation, I think you can contact most one of three or four, okay?

好的。所以我覺得，如今由於對自動駕駛計程車的討論越來越多，技術也越來越成熟，你會看到很多汽車製造商和平台——叫車平台開始談論自動駕駛計程車，並開始公佈他們當地的稅收收費策略。但我想提一點，就像自動駕駛計程車一樣——做自動駕駛計程車並不容易，好嗎？自動駕駛計程車行業需要多年的努力、技術累積和監管探索，這就是為什麼世界上成熟的自動駕駛計程車公司如此之少的原因。如果你接觸到成熟的自動駕駛出租車服務，其中一家已經向公眾開放了無人駕駛出租車運營，我想你可以聯繫到三四家中的一家，好嗎？

So not for many. So it's not because of while you see a few companies getting mature, then then you can announce your strategy, better you need to show whether you have enough technology accumulation, enough experience to do robotaxi. So -- but our realized competitive advantage is still in several areas, first of all, technology, right? WeRide, with our ability to massively deploy both L4 and L2 two-plus level mass production of vehicle help us to actually, first of all, gain data -- to gather data more efficiently and make our algorithms more generalizable. So that one actually in turn strengthens our technology.

所以對很多人來說並非如此。所以，並不是說當你看到一些公司發展成熟後，你就可以宣布你的策略，而是你需要證明你是否有足夠的技術累積和經驗來做無人駕駛計程車。所以——但我們已實現的競爭優勢仍然體現在幾個方面，首先是技術，對吧？WeRide 擁有大規模部署 L4 和 L2 兩級以上自動駕駛汽車的能力，這有助於我們首先獲得數據——更有效地收集數據，使我們的演算法更具通用性。因此，這實際上反過來增強了我們的技術。

Second, our capability actually actually for fast iteration is there. And think about cloud OEMs, right? They have -- usually they have a relatively small ADAS system development team. what you can do for L2+ -- L2+ ADAS system is far from what you can achieve in L4 because L4 is a driverless operation. L2 is just like assistant driving system.

其次，我們確實具備快速迭代的能力。想想雲端設備原始設備製造商（OEM）吧？他們通常擁有一個規模相對較小的ADAS系統開發團隊。 L2+等級的ADAS系統與L4等級相比還有很大的差距，因為L4等級是完全無人駕駛的。L2 就像駕駛輔助系統一樣。

For the L2 system, you don't need to take the final responsibility. But for L4 system, you have to be redundant, have to take the responsibility. Over the past nine years, we will have to accumulate lots of experience. We -- that's why we can roll out a robotaxi service. And I haven't seen any other car OEMs or car-hailing platform being able to do so. So that is one of our advantage. And one thing -- the last thing I want to emphasize is about on the core of the company.

對於 L2 系統，您無需承擔最終責任。但對於 L4 系統來說，你必須具備冗餘性，必須承擔責任。過去九年裡，我們將累積大量的經驗。正因如此，我們才能推出無人駕駛計程車服務。我還沒看到其他任何汽車製造商或叫車平台能夠做到這一點。所以這是我們的優勢之一。最後一點——也是我想強調的一點——是關於公司的核心。

The core of WeRide is really the AI technology. WeRide, since since 1, has been an AI company. and we hired so many top tenants and set up our company for a fast iteration in the AI algorithm. I don't think traditional car OEMs or traditional cloud heating platform are capable of this kind of faster iteration. So let's wait and see. But so far, I haven't seen any major car OEMs or car-handling platform company have successfully rolled out any robotaxi service, driverless robotaxi service to public, okay? That's my answer to these two questions.

WeRide的核心其實是人工智慧技術。WeRide自成立以來一直是一家人工智慧公司。我們聘請了許多頂尖人才，並為人工智慧演算法的快速迭代奠定了基礎。我認為傳統的汽車OEM廠商或傳統的雲端加熱平台無法實現這種更快的迭代速度。讓我們拭目以待。但到目前為止，我還沒有看到任何大型汽車製造商或汽車操控平台公司成功向大眾推出任何無人駕駛計程車服務，懂嗎？這就是我對這兩個問題的回答。

Sorry, one more question from me. So following the last question, do you think the amount of data and the development of AI models have given OEMs certain edge to enter and compete in robotaxi. Is it possible to evolve from L2 to L4?

不好意思，我還有一個問題。那麼，接上一個問題，您認為數據量和人工智慧模型的發展是否為汽車製造商進入自動駕駛計程車領域並與之競爭提供了一定的優勢？L2 有可能進化到 L4 嗎？

Very good question. So first of all, I want us to think about one thing. So who are the best L2++ or ADAS system company in this world? Probably Tesla. In China, I think we can name a few maybe XPeng, Li Auto but if you look at their strategy, they are doing L2++ ADAS system, and they talk about lobotaxi, but when they come to robotaxi, they always try to attack this problem or approach this project directly from L4 level, while there's no L3 strategy, okay? Where is their L-3 strategy. There's no L3 strategy from Tesla. There's no -- Xpeng, Li Auto they all skipped L3. Why is that?

問得好。首先，我想讓我們思考一件事。那麼，世界上最好的L2++或ADAS系統公司是哪家呢？可能是特斯拉。在中國，我想我們可以舉幾個例子，例如小鵬汽車、理想汽車，但如果你看看他們的戰略，他們正在開發 L2++ ADAS 系統，他們談論自動駕駛出租車，但當他們談到自動駕駛出租車時，他們總是試圖直接從 L4 級別解決這個問題或著手這個項目，而沒有 L3 級別的戰略，明白嗎？他們的L-3戰略在哪裡？特斯拉沒有L3級放電策略。沒有——小鵬汽車、理想汽車，他們都跳過了L3。這是為什麼？

Because if you directly grow out to Pat L3 and then to L4, they found it's really very, very difficult. It's just like you're climbing a cliff, in that you may be directly directly solve the problem that is using what your experience direct from L4 system-level problem. That is just like we have already done for the past eight or nine years.

因為如果直接長到 Pat L3，然後再長到 L4，他們發現這真的非常非常困難。這就像你在攀登懸崖一樣，你可能會直接利用你的經驗來解決 L4 系統級問題。這和我們過去八、九年所做的一模一樣。

The technology has been there. We have used deep learning algorithms based on large language intermodal or lots of data, but all of this, I want to say, it's based on our past eight years' experience currently through current OEM can leverage on the cutting edge, large language model stuff. But there are lots of infrastructure that's relevant, like data simulation and the cloud computing platform. All of this, I don't think the car OEMs have enough accumulation. Still take many years of them to really roll out a simulation platform, to roll out the protocol, to roll out the pipeline, to test driveless robotaxi.

技術早已存在。我們已經使用了基於大型語言跨模態或大量資料的深度學習演算法，但我想說的是，所有這些都是基於我們過去八年的經驗，目前透過目前的 OEM 可以利用尖端的大型語言模型技術。但還有很多相關的基礎設施，例如資料模擬和雲端運算平台。我認為，汽車製造商在這方面累積的資源還不夠。即便如此，他們仍然需要很多年的時間才能真正推出模擬平台、推出協議、推出流程，並測試無人駕駛機器人計程車。

One thing I want to emphasize having a pretty good ADAS system can let you drive for 100 miles without takeover is far from to roll out driveless robotaxi. To roll out to drive this robo taxi, you have to capable of making that car drive better install of more than 10,000 miles. So that is a kind of magnitude of difficulties.

我想強調的一點是，擁有一個相當不錯的 ADAS 系統可以讓車輛行駛 100 英里而無需接管，但這距離推出無人駕駛機器人計程車還很遙遠。要駕駛這款無人駕駛計程車，你必須有能力讓這輛車行駛超過 10,000 英里。這就是困難的嚴重程度。

So it's just like streaming is new pool and then you swim and then across the English channel, that's different. So I think -- I'm not saying -- I'm not going to say it's absolutely not possible to gradually grow from L2++ to L4, but it will take a long, long time before than that, I think our first class, the Tier 1 robotaxi company we will, we will have already taken over the global market, have already been very profitable. So time left for this major car OEMs to gradually grow from L2++ to L4 is very, very limited.

這就好比串流媒體是一個新的游泳池，然後你游泳，再橫渡英吉利海峽，那就完全不一樣了。所以我認為——我不是說——我不會說從 L2++ 逐步發展到 L4 絕對不可能，但這需要很長時間。在那之前，我認為我們第一家一級機器人計程車公司，我們已經佔領了全球市場，並且已經非常獲利了。因此，留給這家主要汽車製造商從 L2++ 逐步發展到 L4 的時間非常非常有限。

Any other questions?

還有其他問題嗎？

Liping Zhao, CICC.

趙麗萍，中金公司。

The first, I want to follow the previous technical question. And this question is for Tony because you are quite confident in maintaining the leadership in the industry. What tools and technology approaches help you stay ahead of the curve? Could you please share more color from a technical perspective? And then I'll have a follow-up.

首先，我想接著之前的技術問題繼續討論。這個問題是問托尼的，因為你對保持行業領先地位非常有信心。哪些工具和技術方法能幫助您保持領先地位？您能否從技術角度更詳細地介紹一下？然後我還會跟進。

Okay. So first of all, just as I have discussed, right, there's only one company in the world to my best knowledge that are capable of doing robotaxi, have already achieved open to public driveless operation, at the same time, supply aided system to mass production car company that WeRide. And so we have a so-called dual flywheel strategy that is we can gather the data. We collect it through our local taxi fleet, all kinds of corner cases, and use that to facilitate our L2++ development. At the same time, we can also get what we have collected from the L2++ system like AI drive based on navigation and make our L4 system more stable and more generalizable.

好的。首先，正如我之前討論過的，據我所知，世界上只有一家公司能夠製造無人駕駛出租車，並且已經實現了向公眾開放的無人駕駛運營，同時還向大規模生產的汽車公司提供輔助系統，這家公司就是 WeRide。因此，我們採用了所謂的雙飛輪策略，即我們可以收集數據。我們透過本地計程車車隊收集各種特殊案例，並利用這些案例來促進我們的 L2++ 開發。同時，我們也可以利用從 L2++ 系統中收集到的基於導航的 AI 駕駛等知識，使我們的 L4 系統更加穩定且更具通用性。

And so by combining these two sources of data, two source of problem, we actually gradually evolve to a super platform that are capable of robotaxi at the same time, so at the same time, with limited hardware limited HD map, navigation map make us to cover the whole global market.

因此，透過結合這兩個資料來源、兩個問題來源，我們實際上逐步發展成為一個超級平台，能夠同時實現無人計程車、有限的高清地圖和導航地圖，從而覆蓋整個全球市場。

So these two parts actually leverage on each other, help us to improve or to iterate our algorithm at a speed cannot be achieved by no single strategy company, okay? So that's one of our advantage. And the other thing is like the globalization. Since we have deployed the fleets in the global market, we can collect the data. And also, we can get heterogeneous source of data from all over the world.

所以這兩個部分實際上是相互促進的，幫助我們以任何單一戰略公司都無法達到的速度改進或迭代我們的演算法，明白嗎？這是我們的優勢之一。另一件事就是全球化。由於我們已經在全球市場部署了車隊，因此我們可以收集數據。此外，我們還可以從世界各地取得異質資料來源。

We can have the data that's collected in a very dry climate in Mid East and sometimes we can get data from very humid tropical area in Singapore and very cold area in Japan and in China. And I can't imagine any other company have this kind of wide spectrum of application area, and with all of this data and also we hired a group of talented engineers, we actually evolve very fast. The other thing I want to emphasize is our Genesis platform. It's actually based on physical AI model and with lots of considerations on a really more mobile physical AI, or physical AI world. And it actually seamlessly integrate with our end-to-end system. This kind of simulation platform give us a big advantage to develop.

我們可以獲得在中東非常乾燥的氣候下收集的數據，有時還可以獲得來自新加坡非常潮濕的熱帶地區以及日本和中國非常寒冷地區的數據。我無法想像還有哪家公司擁有如此廣泛的應用領域，而且憑藉所有這些數據，再加上我們聘請了一群才華橫溢的工程師，我們的發展速度非常快。我還要強調一點，那就是我們的 Genesis 平台。它實際上是基於物理人工智慧模型，並且對更具移動性的物理人工智慧或物理人工智慧世界進行了許多考慮。它與我們的端到端系統實現了無縫整合。這種模擬平台為我們帶來了很大的研發優勢。

And with combination of all of these kind of algorithm and data, I think this massive effective here. We have many, many driverless permits in all over the world in many countries. And we have many car OEMs collaborating with us. We have heterogeneous data collected from different country, different level of autonomy. And by combining all of this, I think we can achieve accelerated development speed, which is much faster than our competitors. That's my answer to your question.

我認為，結合所有這些演算法和數據，就能產生巨大的效果。我們在世界各地許多國家都擁有大量的無人駕駛許可證。我們與許多汽車原始設備製造商 (OEM) 都有合作。我們收集了來自不同國家、不同自治程度的異質資料。我認為，透過結合所有這些因素，我們可以實現比競爭對手更快的開發速度。這就是我對你問題的回答。

Tony, that's very helpful. And my second question is for Jennifer. The Board of the company has authorized USD1 million share repurchase program in May this year. And could you please update what is the status on this program?

托尼，這很有幫助。我的第二個問題是問詹妮弗的。公司董事會已於今年5月批准了一項100萬美元的股票回購計畫。請問這個計畫的最新進展如何？

Thank you, Liping. Regarding the USD100 million share repurchase program, which is authorized by the Board in May, we haven't -- no purchase has been initiated to date. The reason is the pre-proprietary work for our Hong Kong IPO constitute for a close period under the securities regulation, which -- during which the trading was restricted. As a new listed company, we are also required to obtain specific shareholder approval to ratify this program. We are currently preparing the call an Extraordinary General Meeting to segues approval, which will allow the program to proceed.

謝謝你，麗萍。關於董事會於 5 月批准的 1 億美元股票回購計劃，我們至今尚未啟動任何回購。原因是，我們香港IPO的上市前準備工作構成證券法規規定的封閉期，在此期間交易受到限制。作為一家新上市的公司，我們也需要獲得股東的特定批准才能批准該計劃。我們目前正在準備召開特別股東大會以獲得批准，這將使該計劃得以繼續進行。

Operator, we're ready to take the next question.

接線員，我們準備好回答下一個問題了。

Paul Gong, UBS.

Paul Gong，瑞銀集團。

Paul Gong from UBS. I have two questions. The first one is -- regarding the robotaxi revenue contribution, we have noticed that while this is about 7 times year-over-year growth it seems to have a little bit fluctuation compared to the second quarter. Could you please elaborate more on this? And my second question is regarding the European strategy. congratulate for the permit in canteen today. Can you share more on the next step of the company's plan for European expansion?

瑞銀集團的Paul Gong。我有兩個問題。第一個問題是——關於自動駕駛計程車的收入貢獻，我們注意到，雖然這比去年同期增長了大約 7 倍，但與第二季度相比似乎略有波動。能詳細解釋一下嗎？我的第二個問題是關於歐洲戰略的。恭喜你今天在食堂獲得了許可證。能否詳細介紹一下公司下一步的歐洲擴張計畫？

Okay, I'll take the first question. I think Tony probably want to take the second. So regarding the fluctuation of robotaxi revenue, we we'll say, for the past three quarters, you can see our robotaxi revenue made like a 20% -- 22% contribution in first quarter, 36% and 21% revenue contribution for the past two quarters, which already showcase a very continuous momentum. The fluctuation was expected, given that our delivery schedule is in tandem with the permit upgrade and the corresponding expansion of our operating area. We made a significant step forward by securing the city level driverless operation permit in Abu Dhabi, which will pave the way for accelerating expansion in the entire Middle East going forward.

好的，我先回答第一個問題。我認為托尼可能想選第二個。關於自動駕駛計程車收入的波動，我們想說，在過去的三個季度中，你可以看到我們的自動駕駛計程車收入在第一季度貢獻了 20% 到 22% 的收入，在過去的兩個季度分別貢獻了 36% 和 21% 的收入，這已經顯示出非常持續的成長勢頭。考慮到我們的交付計劃與許可證升級和我們營運區域的相應擴大同步進行，這種波動是可以預期的。我們在阿布達比獲得了城市級無人駕駛營運許可證，這標誌著我們向前邁出了重要一步，這將為未來在整個中東地區加速擴張鋪平道路。

Yes. Tony, do you want to take the second one?

是的。東尼，你想選第二個嗎？

Yes, I will take the second one. Okay. The second question is about -- actually our -- since we have already got the driverless robotaxi (inaudible) Switzerland and what's our next step for Europe, okay? So European market is a great market for robotaxi. I think the taxi farer in Europe is high and also European -- most of the countries in European market, they are short of labor.

是的，我要第二個。好的。第二個問題是關於——實際上是關於我們——既然我們已經在瑞士實現了無人駕駛機器人出租車（聽不清楚），那麼我們下一步在歐洲的發展方向是什麼，好嗎？因此，歐洲市場對無人駕駛計程車來說是一個非常好的市場。我認為歐洲的計程車費用很高，而且歐洲大部分國家都面臨勞動力短缺的問題。

And that is actually very, very good scenario or opportunity for a robotaxi company to deploy more robotaxi service to fill the gap between the shortage of drivers and increasing demand for taxi these days, we know the numbers that people -- especially after the coronavirus, if possible, people like to take kind of private transportation, if possible. So I think there's an actual growing demand for taxi.

實際上，這對一家機器人計程車公司來說是一個非常非常好的場景或機會，可以部署更多的機器人計程車服務來填補司機短缺和計程車需求不斷增長之間的差距。我們知道，人們——尤其是在新冠疫情之後——如果可能的話，都喜歡乘坐私人交通工具。所以我認為計程車的需求確實在增加。

So if we can deploy our robotaxi as a very cost-effective message, people will love this kind of product. So since Switzerland we have already get it, and now we are considering some other countries like we established an office in Stuttgart and also try to explore in Paris. So in the next 12 months, we will solidify our foundation actually with our trial operation in France, in Belgium and also our current operation in Switzerland and talk to all the possible countries. And we also formed a strategic partnership with Uber, with Renault, with SBB, et cetera. And there are a lot of time for airports in European countries, talking to us to explore the possibility of our robobus.

因此，如果我們能以非常經濟高效的方式推廣我們的無人駕駛計程車，人們就會喜歡這種產品。既然我們已經在瑞士取得了成功，現在我們正在考慮其他國家，例如我們在斯圖加特設立了辦事處，並且也在嘗試探索巴黎市場。因此，在接下來的 12 個月裡，我們將透過在法國、比利時的試營運以及我們在瑞士的現有營運來鞏固我們的基礎，並與所有可能的國家進行洽談。我們也與 Uber、雷諾、瑞士聯邦鐵路公司等建立了策略合作關係。歐洲各國的機場有許多時間與我們洽談，探討我們無人巴士的可能性。

So we want to use all of our applications like robobus, robosweeper to actually help us to explore possibility to to extend to certain countries because in certain European countries, they would like to tend to adopt robobus or robosweeoer first and then try to do local taxi. And gradually, we want to go to Switzerland, Germany, from Spain and Norway, these kind of countries to extend to extend to more countries. And then, of course, I want to emphasize our approach is dynamic.

因此，我們希望利用我們所有的應用程序，如 robobus、robosweeper，來幫助我們探索擴展到某些國家的可能性，因為在某些歐洲國家，他們傾向於先採用 robobus 或 robosweeper，然後再嘗試開展本地出租車業務。然後，我們希望逐步從西班牙和挪威等國家擴展到瑞士、德國等更多國家。當然，我還想強調我們的方法是一個動態的過程。

It depends on the availability of strong local partner and also depends on regulatory policy and the local -- shortage of the labor, all of these factors we have to consider. But for sure, we will gradually expand to the aforementioned countries and all other potential countries in Europe.

這取決於是否有實力雄厚的當地合作夥伴，也取決於監管政策和當地勞動力短缺情況，我們必須考慮所有這些因素。但可以肯定的是，我們將逐步擴展到上述國家以及歐洲所有其他潛在國家。

As there are no further questions, I will conclude the call today. Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.

由於沒有其他問題，今天的通話就到此結束。感謝您參加今天的會議。節目到此結束。您現在可以斷開連線了。

Thank you very much. Thank you. Bye.

非常感謝。謝謝。再見。

Thank you.

謝謝。