格隆汇11月25日｜小马智行(PONY.US)盘前拉升，涨超5.5%，报13.29美元。消息面上，小马智行发布 2025年第三季度财报，总营收1.81亿元，同比增长72%，连续三个季度实现营收增长；Robotaxi业务收 入达4770万元，同比增长89.5%，其中乘客车费收入同比上涨超200%。第七代Robotaxi在广州实现单车 盈利转正，公司将提前完成Robotaxi千辆规模目标，2026年底将扩至3000辆以上。目前公司Robotaxi车 队规模达961辆，其中667辆为第七代Robotaxi，通过引入轻资产模式，加快车队规模化投放。(格隆汇) ...

Core Insights - The article discusses the launch of MiMo-Embodied, the world's first unified base model for autonomous driving and embodied operations, developed by Xiaomi's Chen Long team [1][3]. Group 1: Model Overview - MiMo-Embodied is based on the MiMo-VL architecture and addresses the knowledge transfer challenges between autonomous driving and embodied operation scenarios by creating a high-quality dataset that includes general vision, embodied tasks, and driving scenes [3][10]. - The model employs a progressive four-stage training strategy that incorporates Chain of Thought (CoT) and Reinforcement Learning (RL), achieving state-of-the-art (SOTA) performance across 29 benchmarks in both autonomous driving and embodied intelligence [3][21]. Group 2: Challenges Addressed - Previous models in the embodied and autonomous driving fields lacked a unified embodied Visual Language Model (VLM), which limited their ability to interact effectively with the physical world in dynamic environments [6][9]. - The significant domain gap between indoor operations and outdoor driving has hindered the transfer of capabilities across these two areas [8][10]. Group 3: Training Strategy - The training data encompasses three dimensions: general multimodal understanding, embodied AI (including affordance prediction, planning, and spatial understanding), and autonomous driving (covering perception, prediction, and planning) [15][19]. - The four-stage training strategy includes: 1. **Stage 1**: Embodied AI Supervised Fine-tuning with general and embodied data [18]. 2. **Stage 2**: Autonomous Driving Supervised Fine-tuning, focusing on multi-view spatial reasoning and complex traffic scene analysis [20]. 3. **Stage 3**: CoT Supervised Fine-tuning, enhancing the model's ability to handle complex multi-step problems [20]. 4. **Stage 4**: RL Fine-Tuning using the GRPO algorithm to optimize accuracy and reliability [20]. Group 4: Performance Evaluation - MiMo-Embodied was evaluated through both qualitative and quantitative assessments, demonstrating competitive results against existing models in various benchmarks for embodied intelligence and autonomous driving [21][23]. - In embodied capabilities, MiMo-Embodied showed particular advantages in affordance prediction and spatial understanding compared to other models [23][24]. - The model also excelled in autonomous driving capabilities, showcasing strong performance in perception, prediction, and planning across diverse real-world driving scenarios [25][26]. Group 5: Real-World Applications - In embodied navigation tasks, MiMo-Embodied outperformed models like GPT-4o and Qwen2.5-VL in object localization and consistent performance across varied household scenarios [27]. - The model demonstrated robust affordance and spatial reasoning abilities in operational tasks [29]. - In autonomous driving, MiMo-Embodied effectively handled complex tasks such as turning at intersections and lane changes, integrating road context and vehicle state for coherent decision-making [33][36].

Core Viewpoint - The company reported significant revenue growth driven by its autonomous driving services and technology licensing, alongside a successful dual primary listing in Hong Kong, which will support its commercialization efforts and expansion plans [1][2]. Financial Performance - For Q3 2025, total revenue reached $25.4 million (RMB 181 million), a 72.0% increase from $14.8 million in Q3 2024, primarily due to strong growth in autonomous driving services and technology licensing [1]. - The gross margin improved to 18.4% in Q3 2025 from 9.2% in Q3 2024, attributed to an optimized revenue structure with higher-margin autonomous driving service contributions [1]. - The net loss for the quarter was $61.6 million (RMB 438 million), with a basic loss per share of $0.16 [1]. Operational Highlights - The company achieved single-vehicle profitability for its seventh-generation autonomous taxi within city limits and anticipates exceeding its fleet target of 3,000 vehicles by the end of next year [1]. - The seventh-generation autonomous taxis have commenced full driverless commercial operations in major cities such as Guangzhou, Shenzhen, and Beijing [1]. - Revenue from autonomous driving services grew by 89.5% year-on-year, with passenger fare revenue surging over 200% [1]. Strategic Initiatives - The dual primary listing in Hong Kong raised over $800 million, aimed at accelerating mass production, commercialization, technology R&D, and market expansion [1]. - The company plans to leverage its technological advantages, including the world model and AI-driven self-learning mechanisms, to enhance its competitive edge and adapt to new markets and vehicle platforms [2]. - The recent IPO proceeds are expected to strengthen the company's balance sheet and support rapid expansion of its business footprint for long-term growth [2].

Industry Information - CATL and Haibosich signed a 10-year strategic cooperation agreement to enhance their advantages in energy storage technology and smart manufacturing [9] - The price of lithium iron phosphate has rebounded by approximately 10% since October, driven by strong demand from automakers and battery manufacturers [9] - A breakthrough in the certification and review technology system for domestic automotive chips has been achieved, addressing the challenges of "dare not use, cannot use" [9][10] - As of the end of October, China's electric vehicle charging infrastructure reached 18.645 million units, a year-on-year increase of 54% [11] - A total investment of nearly 10 billion yuan has been signed for 16 key projects in the new energy industry in Changzhou [12] - In the first ten months, Beijing's new energy vehicle production increased by 150% year-on-year [12] - The new AITO M7 has surpassed 30,000 deliveries within 57 days of its launch [12] - Guoxuan High-Tech has officially started mass production of standard battery cells [12] - Huawei's advanced driving solution for L3 level is set to launch commercially, marking a significant step in autonomous driving technology [12][14] - In Liuzhou, the total vehicle sales reached 1.265 million units in the first three quarters, a year-on-year increase of 33.7% [15] Policy Information - Hebei Province has adjusted its time-of-use electricity pricing policy, optimizing charging costs for electric vehicles with a peak-to-valley price difference of 70% [16][18] - Chengdu plans to build over 500,000 electric vehicle charging piles and more than 200 battery swap stations by 2030 [21] - New regulations for electric vehicle charging infrastructure in Guangxi have been proposed to clarify responsibilities among stakeholders [21] - Beijing is increasing financial support for the purchase of new energy vehicles [21] - Shenzhen is providing subsidies of up to 5 million yuan for the construction of solar-storage charging and battery swap projects [21] - The Ministry of Industry and Information Technology has announced mandatory national standards for intelligent connected vehicle safety, marking a key step in autonomous driving regulation [21][25] - The Ministry of Finance and the Ministry of Industry and Information Technology are seeking public opinions on government procurement standards for new energy vehicles [21][26] Company Information - Leap Motor reported a total revenue of 31.562 billion yuan in Q3, a year-on-year increase of 9.1% [32] - Leap Motor achieved a net profit of 150 million yuan in Q3, delivering 173,900 vehicles [32] - Xiaomi's Ultra Club racing service has expanded to its 13th city, adding the Guangdong International Circuit [32] - Xiaomi's HAD autonomous driving system is set to upgrade to a new generation XLA architecture in Q4 [32] - Avita Technology signed a business cooperation framework agreement with the Chongqing branch of the Industrial and Commercial Bank of China [32] - NIO's Firefly right-hand drive model has entered mass production, with the first batch shipping to Singapore [34] - NIO has completed the battery swap route in the western Sichuan region, covering popular tourist areas [34] - BYD's sales in the UK have surpassed 50,000 units [34] - Leap Motor's first overseas model cooperation project with FAW has been launched, expected to go into production next year [34] - Xiaopeng Motors plans to launch seven new models next year, including three super electric range-extended products in Q1 [34] - Seres is actively expanding into overseas markets, focusing on high-end smart electric vehicles [34] - Xiaomi's third-quarter profit from electric vehicles and AI reached 700 million yuan [34] - GAC Aion has become the first company in the country to obtain a high-speed L3 testing license for vehicles in collaboration with Huawei [34] - NIO's ES6 model has reached a cumulative delivery of 300,000 units [34] - Geely and Renault have officially launched their strategic cooperation in Brazil, investing 3.8 billion reais (approximately 5.1 billion yuan) [34] - Xiaomi is expected to complete its annual delivery target of 350,000 vehicles ahead of schedule [34] - Tesla's global supercharging stations have exceeded 75,000 [34] - Xiaomi's 500,000th vehicle has officially rolled off the production line, marking a significant milestone [34] - Li Auto is accelerating product iteration and global expansion, moving away from a "cookie-cutter" design approach [34]

北京经开区2025短视频大赛作品火热征集中 作为北京国际科技创新中心"三城一区"主平台、首都高精尖产业主阵地，北京经开区在人工智能、自动 驾驶、商业航天等领域的创新成果，正通过镜头向世界展现。此次短视频大赛以"潮动亦城"为主题，既 是对"北京范儿"短视频大赛的积极响应，更是经开区展示新质生产力发展成果的重要窗口。大赛旨在通 过全方位、多角度的短视频创作，呈现亦庄新城作为国际一流高端产业综合新城的独特魅力，彰显北京 国际科技创新中心建设中的"亦庄担当"。 当航天梦想照进现实，当智慧出行融入日常，当机器人成为生活伙伴。在北京经开区，"未来"从遥远的 想象变成了触手可及的实景。今年8月，由北京经开区工委宣传文化部、北京经开区文联联合主办，环 球网提供媒体支持的第三届"亦创未来"影像大赛——2025"潮动亦城"短视频大赛正式启动征集，以镜头 为媒，向全网展现亦庄新城的科技活力与青春风貌。 当前，大赛作品正在火热征集中，时间截至2026年2月28日。无论是记录产业创新的硬核瞬间，还是捕 捉新城风貌的人文细节，抑或是展现市民生活的幸福图景，均可参与征集。诚邀广大影像爱好者、科创 从业者、市民游客拿起手机，用镜头定格亦庄的 ...

Investment Rating - The report maintains a "Recommended" investment rating for the automotive industry [6] Core Insights - Current demand is weak, but the annual cumulative figures remain above expectations. Retail sales of passenger cars from November 1-16 reached 886,000 units, a year-on-year decrease of 14%, while cumulative retail sales for the year reached 20.142 million units, a year-on-year increase of 7% [1][21] - The new energy vehicle market shows positive growth, with retail sales of 554,000 units from November 1-16, a year-on-year increase of 2%, and cumulative retail sales for the year reaching 10.703 million units, a year-on-year increase of 21% [1][21] - Multiple companies are accelerating their autonomous driving strategies, with significant advancements in L3 commercial applications. GAC Group has become the first company in China to receive a testing license for L3 autonomous driving at speeds up to 120 km/h [3][43] Summary by Sections Market Overview - The automotive sector experienced a decline of 4.89% in the past week, with most related sub-sectors also declining. The Shanghai Composite Index fell by 3.77% during the same period [12][13] Data Tracking - Passenger car retail and wholesale sales from November 1-16 showed a retail figure of 886,000 units and a wholesale figure of 1.021 million units, both reflecting a year-on-year decrease of 14% [21] - New energy vehicle retail sales for the same period reached 554,000 units, with a cumulative total of 10.703 million units for the year, marking a 21% increase [21] Industry News - Companies like Leap Motor and NIO are making strides in global markets, with Leap Motor's collaboration with FAW Group expected to yield an overseas model by next year [2][38] - GAC Group's A800 has received approval for L3 testing, indicating a significant step towards practical application of autonomous driving technology [3][43] - Xiaomi is set to upgrade its HAD system to a new architecture by Q4 2025, enhancing its autonomous driving capabilities [33][42]

Core Insights - RoboSense has partnered with Didi Autonomous Driving to develop a next-generation Robotaxi model, utilizing its advanced digital lidar solutions for comprehensive perception capabilities [1][2] Group 1: Partnership and Collaboration - RoboSense will provide Didi Autonomous Driving with its EM4 and E1 lidar systems, enabling "long-distance precise identification + near-field blind spot elimination" for the Robotaxi [1] - The collaboration aims to enhance the safety and efficiency of autonomous driving through advanced perception technology [1] Group 2: Technology and Product Features - The EM4 lidar is noted for having the highest production line count and strongest ranging capability in the industry, allowing for accurate detection of distant small obstacles [2] - The E1 lidar features a 120°×90° ultra-wide field of view, effectively eliminating near-field blind spots, making it the preferred solution for Robotaxi applications [2] Group 3: Market Position and Clientele - RoboSense has established partnerships with over 90% of leading Level 4 (L4) autonomous vehicle companies globally, indicating a strong market presence [2] - The EM4 and E1 systems have been selected by multiple top-tier Robotaxi clients in North America and China, with plans to expand their operational fleets by the end of the year [2]

Core Insights - RoboSense has partnered with Didi Autonomous Driving to develop a new generation of Robotaxi models, utilizing advanced digital lidar technology for enhanced perception capabilities [1] Group 1: Partnership and Technology - RoboSense will provide Didi Autonomous Driving with its EM4 and E1 lidar systems, enabling comprehensive perception through "long-distance precise identification + near-field blind spot elimination" [1] - The main configuration for the new Robotaxi perception solution consists of "4 EM4 and 4 to 6 E1 units," showcasing the industry's only mass-producible high-performance lidar technology [1] Group 2: Product Features and Market Position - The EM4 is recognized as the highest production line count and strongest ranging digital lidar in the industry, capable of accurately identifying small obstacles at a distance, thus providing vehicles with more reaction time [2] - The E1, a solid-state blind spot lidar, features a 120°×90° ultra-wide field of view, effectively eliminating near-field blind spots and becoming the preferred solution for Robotaxi deployment [2] - RoboSense has established partnerships with over 90% of leading Level 4 (L4) companies globally, with plans to expand the operational fleet of its North American Robotaxi and Robotruck clients by the end of the year [2]

Core Insights - WeRide, known as the "first Robotaxi stock globally," successfully completed its dual listing on the Hong Kong Stock Exchange, filling a gap in the market for Robotaxi stocks in Hong Kong [1][4] - The company reported strong financial results for Q3 2025, with revenue reaching 171 million yuan, a year-on-year increase of 144.3%, and a significant growth in Robotaxi business revenue [3][5] - The company has established a robust presence in the autonomous driving market, with over 1,600 autonomous vehicles globally, including nearly 750 Robotaxis [3][8] Financial Performance - Q3 2025 revenue was 171 million yuan, marking a 144.3% increase year-on-year, the largest quarterly growth since the company's inception [3] - Robotaxi business revenue reached 35.3 million yuan, a staggering 761.0% increase year-on-year [3] - Gross profit was 56.3 million yuan, up 1,123.9% year-on-year, with a gross margin of 32.9%, an increase of 26.4 percentage points year-on-year [3] Market Position and Strategy - WeRide has secured a unique position as the only technology company with autonomous driving licenses in eight countries, showcasing significant first-mover advantages [8][14] - The company has established partnerships with major players like Uber and Grab, enhancing its market reach and operational capabilities [15] - WeRide's strategic focus on commercializing its autonomous driving technology across various applications, including Robotaxi, Robobus, and Robovan, demonstrates its commitment to diversifying revenue streams [5][17] Technological Advancements - The WeRide One autonomous driving platform and WeRide GENESIS simulation engine create high barriers to entry, ensuring reliability and safety in autonomous driving systems [6][9] - The company has maintained a leading position in safety and reliability, with over 55 million kilometers of public road testing and no regulatory penalties due to system failures [11] Global Expansion - WeRide has expanded its operations to 30 cities across 11 countries, transitioning from pilot operations to large-scale deployment [8][17] - The company has recently received approval for fully autonomous Robotaxi operations in Abu Dhabi, marking a significant milestone in its global expansion strategy [15] - The Robotaxi service is set to launch in Singapore, further solidifying WeRide's presence in key international markets [12][14] Future Outlook - The global Robotaxi service market is projected to reach $66.6 billion by 2030, with a compound annual growth rate of 196% from 2025 to 2030, positioning WeRide to benefit from significant market growth [16] - The company is expected to demonstrate the feasibility and scalability of its autonomous driving business model, paving the way for future financial performance improvements [17][18]

Core Viewpoint - 小马智行-W (2026.HK) has experienced a significant stock price increase of 10.5%, reaching HKD 97.3, with a total market capitalization of HKD 42.1 billion following the announcement of strategic partnerships [1][3]. Group 1: Strategic Partnerships - 小马智行 has entered into a strategic cooperation with the ride-hailing platform 阳光出行 to develop a large-scale autonomous driving fleet based on its seventh-generation production model [3]. - The initial fleet is set to launch this year in major Chinese cities, including Beijing, Guangzhou, Shenzhen, and Shanghai, with plans for further operational integration [3]. - Additionally, 小马智行 has partnered with the Estonian ride-hailing company Bolt to introduce autonomous vehicles into taxi services in Europe [3].