李想2026年2月9日微博： 全新L9用的马赫100双芯片，总算力2560 TOPS。关键不只是总算力，更要看单颗有效算力：马赫100 单颗算力1280 TOPS，因为采用数据流架构提供给算法软件最大的优化空间，单颗马赫100的有效算 力就是英伟达Thor U的3倍，全新L9的双马赫100芯片，有效算力就是Thor U的5-6倍了。 什么是有效算力？就是实际跑VLA大模型时，能榨出来的真实性能。传统GPU架构利用率低，我们 的数据流架构利用率高、功耗低。 具体表现：同样场景下，我们能做到更高帧率、更短反应时间。紧急情况下，可以更早感知风险，更 快做出避险动作——这就是体验代差，不仅是参数好看，是关键时刻管用。 2022年启动自研时，我们就判断：2025年开始，行业进入【自研算法+自研算力】软硬一体时代。马 赫100，是这个判断的第一步。 2025年7月22日36氪汽车发文《 市值第一英伟达，被中国汽车浇冷水｜深氪 》 7月23日进行了部分修改，修改后的内容为： "双方对Thor进行了数轮调整后，才达到量产交付效果。但英伟达原先宣传的700TOPS算力，实际目 前可释放的只在500左右。" 2025年7月27日 ...

Core Viewpoint - The article emphasizes the transition from 2D Vision Transformers (ViT) to 3D ViT, highlighting the advantages of processing continuous video streams for better understanding of the physical world and improved response times in autonomous vehicles [1][2]. Group 1: Transition to 3D ViT - The traditional 2D ViT processes images in a sliced manner, limiting the information captured from each frame, while 3D ViT processes video clips, integrating spatial and temporal data for enhanced feature extraction [1][2]. - The shift to 3D ViT is not merely a conversion from 2D to 3D in terms of perspective but involves a fundamental change in feature extraction dimensions, focusing on height, width, and time [2]. Group 2: Technical Advancements - The new chip architecture, referred to as a data flow architecture, allows for direct connections between layers on the silicon chip, minimizing the need for external memory reads and writes, thus optimizing latency [2]. - The self-developed chip by the company is designed to be data-driven rather than instruction-driven, achieving higher parallelism and integrating hardware and software design from the outset [4]. Group 3: Implications for Autonomous Vehicles - The advancements in chip technology necessitate corresponding improvements in vehicle control systems, leading to the development of a fully controlled chassis for the L9 model to match the enhanced processing capabilities [3].

视频版: 文字版： 李想: 我觉得大家比较好奇的一点，这一次除了我们在讲汽车要实现具身智能化以外，然后我们还开 始把过去隐藏了一年多的时间的人形机器人正式地浮出水面了。 我背后所有的思考其实是这样的，我觉得就是人类所有的我们今天提供的服务、产品、应用，我自己 大概就分了四个大类。 面向企业内的大概有两类，然后面向终端消费者的其实有两类。 面对企业内的两类，一个是信息化工具，一个是生产力工具。当有了AI，很多企业都能够自己非常 好地解决信息工具的问题。 然后另外一个角度，其实从过去我们使用Photoshop、Python，变成了今天的Lovart，Claude Code，真 正用于生产的时候，大家就都会普遍去使用全世界最好的生产工具。 只是今天这些生产工具逐步地AI化了。那在To C的方面，就面向终端消费者方面，有两个大的品 类。 一个品类是如何帮用户解决生活问题的，包含衣食住行，包含所有的应用、服务、软件。我觉得另外 一方面就是娱乐，如何有情绪价值，像游戏，像短视频，像电影。如果回归到理想汽车，我们到底应 该怎么定位？ 我认为我们最大的价值其实是做好生活。车也是我们生活中最主要的一个部分，最重要的是能够在物 ...

Core Viewpoint - The new Li Auto L9 represents a significant advancement in embodied intelligence, with a complete overhaul of perception and cognitive processes, transitioning from 2D ViT to 3D ViT technology [1][2]. Group 1: Technological Advancements - The traditional 2D ViT lacks the ability to truly understand the physical world, while the new 3D ViT allows for human-like perception [2]. - The L9 utilizes a self-developed chip, which significantly enhances computational power, being over three times more powerful than previous chips [3]. - The architecture of the new chip allows for flexible computation, enabling the implementation of 3D ViT through a powerful compilation team [3]. Group 2: Control Systems and Performance - The L9 introduces a fully steer-by-wire system, including electric mechanical brakes and four-wheel steering, which enhances response speed and safety [4][5]. - The new active suspension system, powered by four independent motors, combines comfort, handling, and stability, breaking the traditional trade-offs in automotive design [5]. Group 3: Future of Automotive Industry - The competition in the mid-to-high-end automotive market over the next 3 to 5 years will center around embodied intelligence, similar to the evolution from feature phones to smartphones [7]. - The core barriers in the automotive industry will be determined by the integration of chips and models, emphasizing the importance of co-design [7][8]. - The company has been designing its own chips since 2022 and developing foundational models since 2023, anticipating that consumers will recognize the value of these investments through their experience with the L9 [8].

Core Viewpoint - The new generation of Li Auto L9 focuses on creating a significant experiential difference rather than merely upgrading parameters, aiming to build a vehicle that truly understands its users and possesses vitality [1]. Group 1: Product Features - The Li Auto L9 Livis, priced at 559,800 yuan, represents a flagship SUV that embodies the company's vision for the era of embodied intelligence, achieving three groundbreaking advancements [1]. - It features the world's first 800V fully active suspension system under 600,000 yuan, providing over 10,000 N of lift per wheel, eliminating the need for anti-roll bars, and achieving millisecond-level response for an "anti-physical" vehicle posture [1]. - The L9 Livis includes the world's first "complete" all-line control chassis, which integrates steer-by-wire, four-wheel steering, and the first fully electric mechanical brake (EMB), offering intuitive responses akin to a nervous system and evolving from "human-driven" to "human-machine co-driven" [1][2]. Group 2: Technological Advancements - The vehicle is equipped with a powerful brain featuring 2560 TOPS of computing power, utilizing two self-developed 5nm Mach 100 chips, which provide three times the effective computing power of Nvidia Thor-U, along with 360-degree coverage of laser radar [2]. - The new generation of Li Auto L9 will offer a custom dual-tone body with a unique three-layer painting and seamless manual polishing process, merging extreme technology with aesthetic excellence [2]. Group 3: Market Positioning - The all-new Li Auto L9 is set to be released in the second quarter of this year, making it a worthwhile consideration for consumers looking to purchase a full-size flagship SUV [3].

Core Viewpoint - The company envisions the ultimate form of a car as a robot, not just a faster or smarter transportation tool, and has been progressively working towards this goal over the past decade [1]. Group 1: Development Timeline - In October 2018, the company launched the Li ONE, introducing sensory capabilities with microphones, radars, and cameras [1]. - In May 2021, the self-developed perception system was integrated into the 2021 Li ONE model [1]. - In October 2021, the project for a self-developed operating system was initiated [2]. - In June 2022, the Li L9 was released, featuring the first mass production of the self-developed central domain controller XCU [2]. - In November 2022, the project for self-developed chips was initiated [3]. - In March 2023, the project for a self-developed large model was initiated [4]. - In December 2023, the self-developed large model MindGPT was launched with OTA 5.0 [5]. - In April 2024, the Li L6 was released, marking the first mass production of the self-developed operating system Star Ring OS [6]. - In October 2024, the first end-to-end AD large model was delivered [7]. - In March 2025, the Star Ring OS was open-sourced [8]. - In August 2025, the VLA driver large model was integrated into the Li i8 [9]. Group 2: Future Vision - The upcoming Li L9 is positioned as a smart entity, equipped with a complete technology stack that transforms the car from a passive tool into an active partner [9]. - The company emphasizes that embodied intelligence must be integrated into a good car to create real value for users, dedicating 70% of its focus on automotive development [9]. - The company is preparing for its second decade in 2026, with the new Li L9 being a pioneering product in the realm of embodied intelligent robots [9].

Sales - The sales performance of the 2026 L9 model will significantly impact the revenue, net profit, and free cash flow limits for the company in 2026 and 2027, while the sales of the i6/L6 models will greatly influence the lower limits [1] - Observing delivery volumes, order messages, and consumer feedback one month, three months, and six months after the release of the 2026 L9 will provide better insights into consumer acceptance and the model's market performance [1] Autonomous Driving - The focus should be on the actual user value enhancement of autonomous driving by the company compared to competitors, rather than merely comparing which is superior [2][4] - The company has optimized the response time of its vehicles to approximately 350 milliseconds, which is a significant improvement over the current average of 550 milliseconds, potentially reducing accident rates by 50% [2] User Value and Interaction - The improvement in real user value is a fundamental anchor for sales, while the competition ranking is merely a marketing gimmick [4] - The company is expected to explore new interaction methods for its cockpit, potentially inspired by the "Doubao" smartphone, which emphasizes user-friendly interfaces and efficient operation paths [6][7][8] Embodied Intelligence - The company anticipates that 2026 will mark the beginning of a new era in embodied intelligence within the automotive sector, with products that differ significantly from past offerings [10] - The usability and delivery timeline of the embodied intelligence features in the 2026 L9 model are critical observation points, as initial public sentiment may suggest competitors could easily replicate these innovations [11] Battery and Charging Infrastructure - The 2026 L9 is expected to feature a 70 kWh battery with a CLTC range exceeding 400 kilometers, along with a commitment to a robust charging infrastructure [12] - The company's strategy to integrate charging as a product operation will likely resonate well with a significant portion of pure electric vehicle users, especially given the anticipated lack of comparable competition in 2026 [12]

Group 1 - The core viewpoint is that while AI agents were initially thought to reduce the capability gap between individuals, they are actually amplifying disparities significantly, potentially by factors of ten or a hundred [1] - Current consumer demand and available time are lagging behind productivity growth, leading to a situation where productivity increases may result in internal competition and deflation for a period [1] Group 2 - In 2025, three groundbreaking AI products were identified: Claude Code, Doubao Phone, and Manus, indicating significant advancements in the AI sector [3] - In January 2026, three more remarkable AI products emerged: OpenClaw, MoltBook, and Chrome Gemini, showcasing the rapid evolution of AI technology [3]

标准聚焦日常使用便利性和碰撞救援安全性，体现在三大方面： 2026年2月3日，理想官方公众号理想技术发文 理想汽车首个主导的强制性国家标准GB 48001—2026 《汽车车门把手安全技术要求》正式发布 以下为原文： 2026年1月28日，工业和信息化部组织制定的强制性国家标准《汽车车门把手安全技术要求》（GB 48001—2026）由国家市场监督管理总局、国家标准化管理委员会批准发布，将于2027年1月1日起开 始实施。 《汽车车门把手安全技术要求》是全球首个门把手安全法规，同时也是理想汽车首个主导的强制性国 家标准、首次在强制性国家标准领域承担核心编制任务，彰显了企业技术研发实力与行业引领作用。 一是研究提出了车门把手的手部操作空间及布置区域要求。 提出了手部操作空间、布置区域等标准 要求， 明确了车门外把手的手部操作空间最小尺寸为60mm×25mm×20mm，重点解决隐藏式外门把 手及内门把手操作不便、难以识别等困扰消费者的问题。 二是明确规范了车门把手的机械释放安全要求。 提出了车门内把手和车门外把手均应具有机械释放 功能，确保不可逆约束装置展开或动力电池热扩散等事件发生后通过机械释放车门把手的单次或 ...

理想座舱产品经理小曹同学营业中回复 最开始提出要做艺术相框应用的其实是想哥本人，还是在i6上市共创会的时候讨论出来的。后 来我们深度思考了大家在车里能和什么类型的内容产生共鸣，精调prompt做了现在的模版出来。 除了现在看到的这些，还有很多在 开发过程中生成不稳定而"牺牲"的创意。但我们还会持续做下去滴，还有很多有趣的模版在路上啦 理想品牌部静静关闭弹射起步补充我还记得他当时大概说的是我们能不能最先把它拿上车，我们自己接的得是行业最好的，审美不能 差，不然没法发给别人。自己执行多难跟用户都没关系，就是需要把它做出来，才会产生价值。 2026年1月27日微博用户只喝美式的喵喵提问，这功能哪位产品经理想出 来的，给他（她）加哦。 部分效果可参考电驱蚊香视频： 另外一个使用场景是"复活"去世宠物。 TOP2短评: 李想投入足够精力的话，提升情绪价值的潜力挺大的，不过精力如何分配是个复杂的管理问题。做出足够强的可以跨端使 用的基座模型是一个非常难啃的硬骨头。 加微信，进群深度交流理想实际经营情况与长期基本面。不是官方群，不是车友群。 ...