• 亿邦公众号
    • APP下载

逃离华为“税” 赛力斯绑着豆包开赌了

作者:邢书博 文章来源:虎嗅APP

2026-06-10 09:57

作者:邢书博 文章来源:虎嗅APP

2026-06-10 09:57

邦小白快读

EN
全文速览

本文核心围绕赛力斯推出去华为化的新品牌赛豆这一事件,梳理了事件背景、行业现状与赛力斯的转型困境,干货信息如下

1.核心事件梳理:赛力斯将旗下屡战屡败的蓝电科技更名重组为赛豆科技,引入重庆地方国资、宁德时代等资本增资66.7亿元,主动让渡控股权,新品牌智能化明确去华为化,座舱采用字节豆包AI系统,智驾自研,字节仅提供技术服务,不持股也不造车。

2.两种智能化路线对比:华为提供全栈软硬件一体化方案,能力覆盖座舱、智驾全链条,成本高能力强;豆包仅提供座舱交互层的大模型API服务,成本低落地快,能力边界清晰。

3.赛力斯的核心困境:依赖华为带来了品牌和销量增长,但增收不增利,自研核心技术不足,此次尝试属于自救的新赌博,结果尚未可知。

本文为新能源汽车品牌的路线选择、品牌建设提供了诸多参考干货,具体内容如下

1.绑定第三方技术服务商的利弊参考:和头部全栈技术服务商合作,可以快速实现品牌跃迁,赛力斯原本自有品牌均价仅8-15万元,和华为合作后问界全系均价达40万元,还省下了百亿级的渠道建设投入,实现连续盈利,但弊端是成本过高,会导致净利率低于行业平均,还会丧失技术主导权。

2.分层市场的路线选择启示:高端品牌可以依托头部全栈方案打市场,低端自有品牌可以选择低成本的智能化方案压缩成本,保障利润空间。

3.品牌建设核心警示:频繁切换技术叙事会消耗品牌资产,缺乏核心产品定义能力会导致用户认知混乱,蓝电三次转型始终销量低迷就是典型案例,品牌必须掌握核心技术才能建立稳定认知。

本文梳理了新能源汽车行业的新变化,整理出从业者关注的机会、风险与可参考模式,干货内容如下

1.市场新机会:低端新能源汽车市场存在低成本智能化升级需求,字节豆包的标准化大模型API输出模式,落地速度快、成本低,从联合开发到上车仅需4个月,适合原有低端品牌快速打上智能化标签,满足消费者对智能座舱交互的基础需求。

2.风险提示:仅更换外部AI方案,不解决自身核心技术缺失、产品定义混乱的问题,无法改变销量低迷的现状;同时互联网快速迭代的思路不符合汽车行业监管要求,OTA过快升级存在合规风险。

3.可参考的运作模式:高端市场绑定头部全栈方案商保品牌销量,低端市场用低成本方案测试自有品牌保利润,同时引入国资和产业资本获得更多独立运作空间,分散经营风险。

本文对新能源整车生产工厂的智能化转型发展提供了不少启示,干货内容如下

1.产品生产设计需求参考:当前新能源汽车市场分层明显,高端市场需要全栈智能化支撑的豪华产品,低端入门市场需要在控制成本的前提下满足消费者对智能座舱交互的基础需求,生产端可针对性调整产能布局,赛力斯新品牌就放在凤凰工厂生产,年内推出新车,正是匹配细分市场需求的布局。

2.商业机会梳理:中小整车工厂可以通过引入地方国资、产业资本增资扩股,获得更多独立运作空间,同时选择成本更低的第三方AI技术服务商,降低智能化转型门槛,快速推出符合市场需求的新产品。

3.转型启示:工厂不能长期依赖外部技术服务商的方案,需要逐步沉淀自有核心技术,尤其是底盘、三电、智驾这些底层核心技术,否则始终无法掌握利润分配的主动权,也很难建立稳定的品牌认知。

本文梳理了汽车智能化服务行业的现状、客户痛点与发展趋势,干货内容如下

1.行业发展趋势:当前汽车智能化服务呈现分层发展的态势,头部玩家可做全栈软硬件一体化方案,覆盖从芯片、OS到智驾座舱的全链条,互联网技术厂商可以从应用层切入,做标准化大模型API输出,聚焦座舱交互场景,走差异化竞争路线。

2.客户核心痛点:不少二线车企依赖头部全栈方案成本高,利润空间被大幅压缩,自身又无力承担全栈自研的高昂成本,有低成本快速实现智能化标签的明确需求;不少原有低端自有品牌缺乏智能化能力,急需低成本的升级方案。

3.服务商发展警示:汽车行业对安全、稳定性、车规级验证要求高,互联网的快速迭代思路并不适配,服务商需要明确自身能力边界,做好隐私合规、稳定性适配,不要盲目切入智驾决策层等核心领域。

本文反映了当前新能源车企对合作平台的新需求,整理了行业新动态与风险规避方向,干货内容如下

1.车企合作需求的新变化:车企不再满足单一头部全栈方案平台的合作,为了降低成本、掌握经营自主权,开始寻求多元化合作,对低成本的应用层AI技术平台有明确需求,希望通过多平台布局平衡成本和自主权的关系。

2.平台运营启示:全栈式汽车技术平台需要持续筑牢全栈自研的护城河,从芯片、OS到智驾算法的全链条整合能力是核心竞争力;应用层技术平台需要做好车规级稳定性和合规性建设,才能持续获得车企客户认可。

3.风险规避提示:汽车行业监管对OTA升级、安全冗余有明确要求,互联网属性的技术平台不能照搬快速迭代的思路,需要符合汽车行业的监管规则和产品要求,避免合规风险,同时要明确自身定位,清晰区分技术服务商和车企的边界,避免市场误解。

本文呈现了中国新能源汽车产业智能化浪潮中的新动向与新问题,对产业研究有较高参考价值,干货内容如下

1.产业新动向:二线主机厂开始探索多元化智能化路线,普遍做法是绑定头部全栈方案商做高端品牌,同时推出独立运作的低端自有品牌,采用低成本的互联网大模型座舱方案,尝试降低对核心供应商的依赖,掌握经营自主权;资本层面,通过引入地方国资和产业资本,主动让渡控股权换取更大运作空间,成为二线主机厂转型的新操作模式。

2.产业新问题:当前不少二线主机厂都存在供应商驱动的问题,缺乏自身产品定义能力和核心技术沉淀,普遍陷入增收不增利的困境,对单一供应商依赖度过高,难以掌握利润分配主动权;频繁切换品牌技术叙事会持续消耗品牌资产,难以建立稳定用户认知。

3.商业模式研究启示:高端绑定全栈方案+低端低成本尝试+多元资本引入的模式,是二线主机厂应对转型的新尝试,但模式能否成功,核心还是要看车企能否完成从供应商驱动到自主定义产品的转型。

声明:快读内容全程由AI生成,请注意甄别信息。如您发现问题,请发送邮件至 run@ebrun.com 。

我是 品牌商 卖家 工厂 服务商 平台商 研究者 帮我再读一遍。

Quick Summary

This article centers on Seres' launch of its new "de-Huaweiized" brand Saidou, and outlines the event's background, the current state of the industry, and Seres' transformation challenges, with key insights as follows:

1. Core event overview: Seres has rebranded and restructured its underperforming Landian brand into Saidou Technology, bringing in Chongqing local state-owned capital and CATL for a total capital increase of 6.67 billion yuan, and proactively ceded controlling stake. The new brand explicitly pursues a "de-Huaweiized" smart vehicle strategy: its cabin uses ByteDance's Doubao AI system, while intelligent driving is developed in-house. ByteDance only provides technical services, holding no stake and not engaging in vehicle manufacturing.

2. Comparison of two smart vehicle development paths: Huawei offers a full-stack integrated hardware and software solution covering the entire smart cabin and intelligent driving chain, with strong capabilities but high costs; Doubao only provides large model API services for the cabin interaction layer, with low costs, fast implementation, and clear capability boundaries.

3. Seres' core dilemma: Dependence on Huawei brought growth in brand awareness and sales volume, but revenue growth did not translate into profits, and the company lacks core in-house technology. This new attempt is a high-stakes gamble for self-rescue, and its outcome remains uncertain.

This article provides valuable insights for new energy vehicle (NEV) brands on path selection and brand building, with key takeaways as follows:

1. Reference for pros and cons of partnering with third-party technical service providers: Partnering with a leading full-stack technology provider allows rapid brand upgrading. The average price of Seres' original in-house brand was only 80,000 to 150,000 yuan; after partnering with Huawei, the average price of its AITO brand reached 400,000 yuan, while Seres also saved tens of billions in channel construction costs and achieved consecutive profits. The downsides, however, are high costs that push net margins below the industry average, and the loss of technological control.

2. Insights for path selection across tiered markets: Premium brands can enter the market relying on leading full-stack solutions, while entry-level in-house brands can adopt low-cost smart solutions to cut costs and protect profit margins.

3. Key warning for brand building: Frequent shifts in technological positioning erodes brand equity, and lack of core product definition capability confuses consumer perception. The consistent low sales of Landian across three transformations is a typical case. Brands must master core technology to build stable consumer perception.

This article sorts out new changes in the NEV industry, and summarizes opportunities, risks, and referenceable models for industry practitioners, with key insights as follows:

1. New market opportunities: The entry-level NEV market has demand for low-cost intelligent upgrading. ByteDance Doubao's standardized large model API output model features fast implementation and low costs, taking only 4 months from joint development to vehicle integration. It is suitable for existing entry-level brands to quickly add an intelligent label and meet consumers' basic demand for smart cabin interaction.

2. Risk warning: Only replacing an external AI solution cannot fix the underlying problems of missing core technology and unclear product definition, and will not reverse low sales. Meanwhile, the rapid iteration approach common in the internet industry does not meet automotive regulatory requirements, and excessively fast OTA updates carry compliance risks.

3. Referenceable operating model: For the premium market, partner with leading full-stack solution providers to guarantee brand sales; for the entry-level market, use low-cost solutions to test in-house brands and protect profits. At the same time, bringing in state-owned and industrial capital creates more space for independent operation and diversifies business risks.

This article offers many insights for the intelligent transformation of NEV manufacturing plants, with key takeaways as follows:

1. Reference for production and design requirements: The current NEV market is clearly segmented: the premium market requires luxury products supported by full-stack intelligence, while the entry-level market needs to meet basic consumer demand for smart cabin interaction while controlling costs. Production facilities can adjust capacity layout accordingly. Seres locates its new brand production at its Fenghuang Plant, with new models launching this year, which is exactly a layout aligned with the needs of the segmented market.

2. Overview of business opportunities: Small and medium-sized vehicle manufacturers can attract capital increases from local state-owned and industrial capital to gain more room for independent operation, while selecting lower-cost third-party AI technology providers lowers the barrier to intelligent transformation, allowing them to quickly launch new products that meet market demand.

3. Transformation insights: Factories cannot rely on external technology provider solutions long-term; they need to gradually build up core in-house technology, especially underlying core technologies such as chassis, three-electric systems, and intelligent driving. Otherwise, they will never hold the initiative in profit distribution, and will struggle to build stable brand perception.

This article sorts out the current state of the automotive intelligent service industry, customer pain points, and development trends, with key insights as follows:

1. Industry development trends: Automotive intelligent services are currently developing along a segmented trajectory. Leading players can build full-stack integrated hardware and software solutions covering the entire chain from chips and OS to intelligent driving and cabins. Internet technology companies can enter from the application layer, output standardized large model APIs focused on cabin interaction scenarios, and pursue a differentiated competition strategy.

2. Core customer pain points: Many second-tier automakers face high costs from relying on leading full-stack solutions, which greatly compresses their profit margins, while they cannot afford the high cost of full-stack in-house development. They have clear demand for low-cost, fast ways to gain an intelligent positioning; many existing entry-level in-house brands also lack intelligent capability and urgently need low-cost upgrading solutions.

3. Warning for service providers: The automotive industry has strict requirements for safety, stability, and automotive-grade validation, and the internet's rapid iteration approach is not a good fit. Service providers need to clarify their capability boundaries, ensure privacy compliance and stability adaptation, and avoid blindly entering core areas such as intelligent driving decision-making.

This article reflects new demands from NEV automakers for partner platforms, and summarizes new industry dynamics and risk mitigation directions, with key insights as follows:

1. New changes in automakers' partnership demand: Automakers are no longer satisfied with partnering with only a single leading full-stack solution platform. To cut costs and gain operational autonomy, they are seeking diversified partnerships, and have clear demand for low-cost application-layer AI technology platforms, hoping to balance cost and autonomy through multi-platform layout.

2. Insights for platform operation: Full-stack automotive technology platforms need to continuously strengthen their full-stack in-house moat; end-to-end integration capability from chips and OS to intelligent driving algorithms is core competitiveness. Application-layer technology platforms need to complete automotive-grade stability and compliance construction to win continuous recognition from automaker clients.

3. Risk mitigation reminders: Automotive industry regulations have clear requirements for OTA upgrades and safety redundancy. Internet-focused technology platforms cannot copy the rapid iteration approach, and must comply with automotive industry regulatory rules and product requirements to avoid compliance risks. They also need to clarify their positioning, clearly draw the line between technology service providers and automakers, and avoid market misunderstanding.

This article presents new trends and issues amid the intelligentization wave of China's NEV industry, and offers high reference value for industrial research, with key insights as follows:

1. New industrial trends: Second-tier automakers are beginning to explore diversified intelligent development paths. The common approach is to partner with leading full-stack solution providers for premium brands, while launching independently operated entry-level in-house brands that adopt low-cost internet large model cabin solutions, in an attempt to reduce dependence on core suppliers and gain operational autonomy. On the capital side, proactively ceding controlling stakes to bring in local state-owned and industrial capital in exchange for greater operating space has become a new transformation model for second-tier automakers.

2. New industrial issues: Many second-tier automakers currently face the problem of being supplier-driven, lacking in-house product definition capability and core technology accumulation. They are generally trapped in the dilemma of revenue growth without profit, over-reliance on a single supplier, and struggle to gain the initiative in profit distribution. Frequent shifts in brand and technological positioning also continuously erodes brand equity, making it difficult to build stable consumer perception.

3. Insights for business model research: The model of "premium brand with full-stack solution + low-cost entry-level attempt + diversified capital introduction" is a new attempt by second-tier automakers to cope with transformation. But the success of the model ultimately depends on whether automakers can complete the transformation from being supplier-driven to independently defining their products.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

赛力斯的灵魂还悬着

出品丨虎嗅汽车组

作者丨邢书博

春秋时期,郑国夹在晋楚两大霸主之间,朝晋暮楚,两头纳贡。郑国大夫子产曾叹道:“小国之事大国也,德,则其人也;不德,则其鹿也。”

这话翻译成白话文,意思是:“小国侍奉大国,如果大国施行仁德,那小国就会像子民一样归附顺从;如果大国不行仁德,那小国就会像被追逐的麋鹿一样仓皇奔逃、叛离而去。”

华为没有“不德”,但赛力斯确实踏入了“鹿”的困境。

众所周知,问界将赛力斯带入了过去从未企及的价格带,也将其推入了一个新问题:当销量、品牌和智能化能力高度依赖华为之后,这家曾经的“西南小厂”还剩多少可独立支配的利润和能力?

赛力斯像“鹿”一样,需要寻找新的生路。

日前,赛力斯旗下原蓝电科技完成更名重组,“赛豆科技”浮出水面。

6月9日赛豆科技发布全新汽车品牌,主打年轻运动市场,首款跨界车型(无实车)提供纯电和增程两种版本,预计年内上市,生产地点为赛力斯凤凰工厂。

赛豆科技的前身,是赛力斯旗下屡战屡败的蓝电科技。2025年5月29日完成更名重组后,通过增资扩股获得66.7亿元注资,股权结构发生根本性调整:重庆沙坪坝区国资旗下的沙磁致远成为第一大股东(持股34.5%),赛力斯退居第二大股东(持股约32.96%),员工持股平台岳行嘉升持股16.48%,宁德时代旗下问鼎投资亦参与入股。

赛力斯主动让出控股权,引入地方国资和产业资本,为的就是给新品牌争取更大的独立运作空间。

值得注意的是,该品牌的智能化方案明确“去华为化”:座舱交互采用字节豆包AI系统,智能驾驶则走自研路线。

正当外界猜测这一新品牌是字节造车前奏时,6月6日,字节跳动通过官方渠道发布声明,澄清了若干市场传言:字节跳动没有造车或推出汽车品牌的计划;“赛豆”不是字节跳动或豆包推出的汽车品牌,字节和赛豆没有股权合作;豆包、火山引擎与汽车行业伙伴的合作,主要是提供豆包大模型、智能座舱等技术服务。

简而言之,字节的角色止步于技术服务商,赛豆汽车的品牌归属、运营主体和资本结构均与字节无关。

换句话说,赛力斯正在问界之外,测试一套成本更低、路径更轻的智能化方案。但这一次,字节只是一个技术供应商,真正的操盘手仍然是赛力斯自己。

通过赛豆科技和自研智驾,赛力斯正在展开一场轰轰烈烈的去华为运动,不过这次能赌对吗?

华为不是“税”,但比较贵

外界常将赛力斯向华为支付的巨额款项称为“华为税”,但这笔账需要算得更清楚。

根据赛力斯招股书及相关公告,2022至2025年上半年,赛力斯向第一大供应商(普遍指华为)的累计采购金额超过750亿元。分年度看,2022年约58亿元,2023年72亿元,2024年跃升至420亿元,2025年上半年约为200亿元。

需要说明的是,这一口径主要为“购买商品及接受劳务”,包含了电驱、域控、传感器等大量实体零部件成本,并不包含市场推广费用,也不包含赛力斯以115亿元入股华为车BU子公司“引望”的投资款。

因此,这750亿远非华为的“纯利润”。在制造业逻辑下,这是包含硬件成本的正常采购。外界是在拿互联网行业的计算方式生拉硬套汽车行业,得出"华为税"的结论很明显是错误的。因为互联网行业没有实体零件成本。

那么,赛力斯用这笔钱换来了什么?

首先是品牌与销量的跃迁。问界M9连续21个月位居50万元级豪华SUV销量冠军,推动赛力斯2025年营收达1650.5亿元,连续两年盈利。而在此前,赛力斯自有品牌均价仅8-15万元。

其次是渠道与能力的补课。问界高度依赖华为全国约700家体验中心,广告及渠道支出从2022年的40亿元飙升至2024年的181亿元。自建同等规模渠道需百亿投入和数年时间。

最后是资本层面的深度绑定。2025年9月,赛力斯公告完成对引望10%股权的收购,创始人张兴海进入引望董事会。双方从业务合作升级为“业务+股权”的战略同盟。

反过来想,若不合作,赛力斯大概率仍停留在蓝电品牌年销不足3万辆、均价12万元的水平,年营收恐难过300亿。

结论是:说华为是“税”是错误认知。恰恰相反,华为是赛力斯最重要的增长引擎,虽然昂贵,但卓有成效。

问题在于,这台引擎越强,越证明赛力斯自己的发动机尚未真正点火。就像问界的增程车一样,华为提供了高价值能力,而赛力斯还想找另一台“增程器”,把自有品牌的续航拉长一点。于是,它找到了豆包。

豆包能做什么,不能做什么?

字节跳动的汽车布局带有鲜明的互联网“铺量”色彩。除蓝电外,豆包大模型已上车长安福特、大众中国、长城汽车、吉利汽车等多家车企,车型从10万级到30万级不等,呈现广撒网态势。

字节的核心优势在于:低成本、快接入、强内容生态。豆包主要提供Agentic AI能力,聚焦智能座舱的语音交互、多轮对话、内容推荐和情绪陪伴。这种标准化API输出模式,对蓝电这类需要快速获得智能化标签、但无力承担完整华为方案的低价位品牌,确实有吸引力。

比亚迪集团高级副总裁杨冬生曾评价豆包,双方从联合开发到上车落地仅用时4个月,体现了高效的“中国速度”。

但豆包的能力边界同样清晰。

第一,它止步于座舱交互层。豆包不涉及座舱OS与车控的深层融合,更不触及智驾决策和安全冗余。华为的鸿蒙座舱是与车机OS、硬件、传感器和应用生态深度绑定的整体方案,能实现“检测到后排睡着后自动降幕、调暗灯光、调低音量”这类需要跨系统联动的主动服务。这种软硬件协同能力,建立在华为数十年的研发投入和车规级验证之上,非豆包短期可及。车载Agent护城河在哪?

第二,字节缺乏汽车工程化经验。 华为车BU成立超过十年,仅智能汽车解决方案业务部就有7000多名研发人员;而字节的汽车团队规模有限。华为做的是从芯片(麒麟)到操作系统(鸿蒙)再到智驾算法的全栈方案;字节做的是大模型API输出,停留在应用层。

有些评论用智驾风险来质疑豆包座舱合作,是不对位的。 如果豆包仅停留在座舱交互层,其核心风险不是“驾驶幻觉致命”,而是车规级稳定性、端云响应延迟、车控权限边界和隐私合规。只有当大模型能力进入智驾决策层,VLA(视觉-语言-行动)架构的幻觉问题才成为核心风险。目前,豆包离这一层还很远。

因此,字节并非在系统性“挖华为墙脚”,而是从座舱交互层切入汽车智能化。它距离挑战华为的核心城池,仍有很长的路要走。

此外,更核心的一点原因是,互联网行业强调的快速迭代和插件化思路,在软硬一体的汽车行业不是优势,反而成了风险点。

从工信部限制车企OTA升级频率,到全行业开启智驾兜底政策看,互联网企业如果还像过去一样先做60分再通过OTA解决缺陷,必然会迎来监管重拳。而用软件定义汽车再做插件,如果不能为智驾、智舱兜底,也会被消费者直接抛弃。

赛力斯蓝电的困境:品牌摇摆与技术空心化

赛力斯将豆包合作放在蓝电品牌上,本身就很说明问题。蓝电几乎是赛力斯“品牌透支”和“战略摇摆”的活教材。

据公开信息,蓝电品牌经历了三套技术叙事:早期脱胎于东风小康体系,首款车型蓝电E5在2023年上市,当年销量仅2.16万辆;面对问界火爆,蓝电E5曾“碰瓷”宣传搭载华为技术,但实际仅使用了HUAWEI HiCar映射方案,销量依然惨淡(2024年全年2.38万辆);如今转向豆包,已是第三次转型。进入2026年,乘联会数据显示,1-4月蓝电品牌累计销量仅1.1万辆,月均不足3000辆。

核心症结在于两点:

第一,产品定义能力缺失。行业分析认为,蓝电E5/E5 PLUS的核心平台架构源于东风风光580,底盘和三电系统无核心竞争力。赛力斯始终在试图通过“外部贴牌”实现溢价,而非像比亚迪那样沉淀核心技术。

第二,品牌资产摇摆。有了华为外围技术时,它试图借华为心智抬高品牌,但实际技术含量不足,反而放大了消费者对“蹭华为”的质疑。如今贴上豆包标签,更像一个大模型厂商的试验田,而非一个成熟的汽车品牌。

过去几次技术叙事切换,都没有真正转化为蓝电稳定的品牌资产。这才是蓝电屡战屡败的根本原因。

赛力斯真正的自救在于定义自己

赛力斯之所以紧急拉拢字节,根源在于其财务状况并不像销量数字那么光鲜。

2026年一季报显示,赛力斯营收356亿元,同比增速放缓至18%,净利润8.7亿元,净利率勉强维持在2.4%左右,低于汽车行业3.2%的平均净利率。而问界终端价格战压力持续,华为技术分成比例依然坚挺。

“增收不增利”的阴影并未散去。本质上,赛力斯需要卖出极高的量级,才能覆盖分给华为的技术授权费、零部件费用和销售提成。

因此,赛力斯的自救逻辑是:高端靠华为保品牌和销量,低端靠豆包保微薄利润。 如果豆包方案能在蓝电上跑通,用低成本座舱体验替代高成本鸿蒙方案,赛力斯就有机会在低端市场达到行业平均利润率,至少能和零跑、五菱坐到一桌。

但这里存在一个根本性悖论:赛力斯自己的灵魂在哪?

真正的自救,应该借问界赚到的现金流,投入到底层核心技术的自研上。2024年,赛力斯研发投入70.53亿元,占总营收的4.86%。但这其中相当部分是对华为智选车方案的适配性开发,真正属于赛力斯自研的纯电平台和智驾算法少之又少。

如果只是把原本给华为的利润,通过压低采购价转移给字节,换来的不过是另一个低成本叙事。蓝电因缺乏核心技术导致的惨淡销量,不会因为语音助手变聪明了就发生质变。

最后的话

平心而论,华为没有亏待赛力斯。

问界M9稳居50万元级车型销量冠军,全系成交均价达40万元,赛力斯连续两年盈利并完成港股上市,成为国内首家“A+H”上市的豪华新能源车企。

豆包合作或许能帮助赛力斯提升自有品牌的智能化体验。但当下,赛力斯的第一要务,仍是紧紧跟随华为,把问界的基本盘做足做好。否则,即便豆包再好用,也暂时撑不起“豪华新能源”这个名头。

其实,赛力斯的处境,是中国二线主机厂在智能电动化浪潮中挣扎的缩影。可以说,没有品牌技术主导权,就没有利润分配权。如果蓝电不能解决产品定义混乱的根本问题,再聪明的大模型也只是一套漂亮的装饰。

字节在城墙外围用大模型和流量攻势不断袭扰,但华为凭借全栈自研的护城河仍扼守着核心城池——赛力斯困在中间。

对于在两个巨头之间行船的张兴海而言,赛力斯的问题从来不是选华为还是选豆包,而是它能否从供应商驱动,转向自己定义产品。否则,即便逃离了所谓“华为税”的围城,也可能迷失在字节的“流量迷宫”里。

注:文/邢书博,文章来源:虎嗅APP(公众号ID:huxiu_com),本文为作者独立观点,不代表亿邦动力立场。

【版权提示】亿邦动力倡导尊重与保护知识产权。未经许可,任何人不得复制、转载、或以其他方式使用本网站的内容。如发现本站文章存在版权问题,烦请提供版权疑问、身份证明、版权证明、联系方式等发邮件至run@ebrun.com,我们将及时沟通与处理。

已点赞

+1

{{ praise_text }}

好文章,值得被更多人看到

广告

下一篇

跨境电商资源对接

《全球跨境电商服务商资源手册》第二期手册现已发布,触达更多资源,了解跨境服务商信息,欢迎点击链接下载。

了解详情

城市微度假火了 锅圈找到第二张「餐桌」

专注家庭餐桌的锅圈,为何进军城市露营场景?这是否预示着露营行业将迎来新的产业整合者?