广告
加载中

硅基流动完成新一轮超20亿元融资 华兴资本担任独家财务顾问

龚作仁 2026-06-17 12:05
龚作仁 2026/06/17 12:05

邦小白快读

EN
全文速览

本文核心信息是第三方MaaS平台硅基流动完成超20亿元B轮融资,获得全产业链头部企业和资本联合投资,其核心业务和发展干货如下

1. 核心业务与市场成绩:硅基流动主打Token工厂模式,当前日均Token调用量达数万亿,服务超1000万用户和1万家企业,营收同比增长超10倍,海外单月营收达数百万美元,多项第三方数据显示其市场影响力位居行业第一梯队,全球Token消耗量排名位居前列。

2. 模式核心价值:Token工厂是AI时代确定性最高的商业机遇,是所有AI应用都依赖的底层基础设施,相当于智能时代的水电煤,IDC预测2026年中国Token消耗量将较2025年增长20倍,市场需求即将爆发,硅基流动已经成为该赛道的领跑者。

本文披露了AI基础设施赛道头部玩家的最新动态,对布局AI转型的品牌商,干货内容总结如下

1. 产业与消费趋势:当前AI已经进入Token经济爆发期,全产业数字化智能化转型加速,AI落地成为必然方向,预计2026年中国Token消耗量增长20倍,底层Token供应缺口大,成本和稳定性是品牌AI落地的核心痛点。

2. 可对接的资源与机会:硅基流动已经适配超160款主流开源模型,支持全品类国产芯片,可提供公有云、私有化部署等多种MaaS服务,能满足不同行业品牌的AI落地需求,品牌商可借助其能力降低自身AI转型的成本,无需从零搭建底层基建。

3. 品牌生态合作空间:硅基流动获得全产业链上下游资本加持,可联动电力、算力、模型、应用等全链路资源,能帮助品牌对接更多生态资源,加速AI商业落地。

对布局AI相关业务的卖家,本文释放的干货机会和提示如下

1. 明确赛道增长机会:Token经济是当前AI领域确定性最高的商业机遇,预计2026年中国Token消耗量较2025年增长20倍,市场需求即将大规模爆发,当前优质Token供给严重不足,赛道存在大量空白机会。

2. 可参考的成熟商业模式:Token工厂作为所有AI应用的底层底座,商业模式清晰,硅基流动成立不到三年就实现营收增长超10倍,海外业务也快速起量,验证了该模式的商业化可行性,卖家可参考该路径切入赛道。

3. 风险与机会提示:当前AI落地的核心风险是底层Token供给不稳定、成本过高,卖家如果做上层AI应用,可以对接成熟第三方MaaS平台,规避底层供应链风险,还能借助其生态资源降低落地成本,抓住AI爆发的红利。

对制造工厂而言,本文在数字化转型和商业机会挖掘层面,干货总结如下

1. AI生产转型的需求参考:当前全产业都在推进AI赋能生产,工厂的数字化智能化转型,对低成本、合规的AI底层能力有大量需求,硅基流动可提供私有化部署的Token工厂,满足国央企、制造企业对数据合规的高要求,能帮助工厂低成本落地AI。

2. 闲置资源的商业机会:如果工厂布局了智算中心或者拥有闲置算力资源,可借鉴硅基流动的解决方案,将裸算力一键转化为标准化Token工厂,解决异构算力适配难、利用率低、资产回报率低的痛点,最大化挖掘算力资产的商业价值。

3. 数字化转型启示:工厂推进AI转型不需要从零投入研发底层AI基建,可借助成熟第三方MaaS平台的能力,降低转型的技术门槛和资金投入,快速将AI应用到生产、管理、销售等各个环节。

对AI相关服务商而言,本文整理的行业干货如下

1. 行业发展趋势:当前AI产业已经迎来Token经济爆发的历史性拐点,Token工厂作为智能时代的底层基础设施,相当于传统经济中的水电煤,市场空间巨大,且商业模式的确定性远高于上层不定向的AI应用,是AI服务商值得布局的黄金赛道。

2. 行业核心客户痛点:当前产业端AI规模化落地的核心痛点,是算力存在结构性紧缺、碎片化、利用率低下的问题,导致优质Token供给严重不足,Token成本过高,制约了AI的落地推广,服务商只要能解决这个痛点就能获得大量市场机会。

3. 成熟解决方案参考:可借鉴硅基流动的路径,通过自研推理引擎做原子级工程优化,极致释放算力效能,再搭建异构算力纳管调度系统,统一调度多元异构算力,提升利用率,快速适配各类主流模型,就能形成可落地的解决方案,满足产业需求。

对AI相关平台商而言,本文的干货内容总结如下

1. 产业对平台的核心需求:当前全产业对大规模、高性价比、稳定的Token供应需求即将爆发,现有供给无法满足市场缺口,AI基础设施平台迎来重大发展机遇,布局Token工厂赛道符合产业发展方向。

2. 平台发展路径参考:硅基流动的发展路径可借鉴,从研发大模型推理引擎切入打好技术基础,再依次上线公有云MaaS、私有化MaaS,覆盖不同类型客户需求,后续推出弹性算力调度引擎优化用户体验,同时对接全产业链资本和生态资源,实现快速增长,路径清晰可复制。

3. 风险规避方向:平台布局该赛道需要提前解决好多元芯片适配、异构算力调度的核心技术问题,同时要兼顾公有云服务和私有化部署能力,满足不同客户的数据合规需求,破解算力碎片化、利用率低的痛点,规避技术和市场风险。

对AI产业研究者而言,本文提供的最新产业研究干货如下

1. 最新产业动向:当前AI产业已经迎来历史性拐点,进入Token经济全面爆发阶段,Token工厂作为AI产业的确定性底层底座,已经成为全产业链布局的热点赛道,头部创业公司硅基流动已经获得能源、芯片、算力、大模型等全产业链头部资本联合投资,商业化增速极快,已经跻身全球头部第三方MaaS平台行列。

2. 行业新问题:当前AI产业发展暴露出多个新问题,包括算力结构性紧缺、碎片化、利用率低下,优质Token供给严重不足,高企的Token成本制约AI规模化落地,这些都是值得深入研究的产业方向。

3. 创新商业模式研究:Token工厂是AI产业分工下诞生的创新商业模式,将算力、算法、系统能力封装为标准化Token输出,作为智能时代的底层基础设施,商业模式清晰,已经验证了商业化可行性,为AI产业专业化分工提供了新的方向,具备很高的研究价值。

返回默认

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 covers the core announcement that Silicon Based Flow, a third-party Model-as-a-Service (MaaS) platform, has completed a Series B funding round exceeding RMB 2 billion, with co-investment from leading players across the AI industry chain. Key takeaways are as follows:

1. Core business and market performance: Silicon Based Flow focuses on the "Token Factory" business model. It currently processes trillions of Token calls daily, serves over 10 million users and 10,000 enterprises, and has achieved over 10x year-over-year revenue growth. Its overseas business generates millions of dollars in monthly revenue. Multiple third-party data sources rank it among the top tier of the industry in market influence, and it holds a leading position in global Token consumption volume.

2. Core value of the model: The Token Factory is the highest-certainty business opportunity in the AI era, serving as core infrastructure that all AI applications rely on — equivalent to electricity, water and gas in the intelligent age. IDC forecasts that China's Token consumption will grow 20-fold from 2025 to 2026, indicating imminent market demand explosion, and Silicon Based Flow has emerged as the front-runner in this track.

This article discloses the latest update of a leading player in the AI infrastructure space, with key takeaways for brands pursuing AI transformation summarized below:

1. Industry and consumer trends: AI has now entered the explosive growth phase of the Token economy, as digital and intelligent transformation across all industries accelerates and AI deployment becomes inevitable. China's Token consumption is projected to grow 20-fold by 2026, creating a large supply gap for upstream Token capacity. Cost and supply stability are the core pain points for brands rolling out AI applications.

2. Accessible resources and opportunities: Silicon Based Flow supports over 160 mainstream open-source models and all categories of domestic AI chips, and offers flexible MaaS solutions including public cloud and private deployment to meet the AI deployment needs of brands across different industries. Brands can leverage its capabilities to cut AI transformation costs, eliminating the need to build underlying infrastructure from scratch.

3. Brand ecosystem collaboration opportunities: Backed by investors across the full AI industry chain, Silicon Based Flow can connect end-to-end resources across power, computing power, models and applications. It can help brands access more ecosystem resources to accelerate the commercial deployment of AI.

Below are key opportunity takeaways and insights for sellers developing AI-related businesses:

1. Clear track growth opportunities: The Token economy is currently the highest-certainty business opportunity in the AI sector. China's Token consumption is projected to grow 20-fold between 2025 and 2026, bringing a massive imminent surge in market demand. There is currently a severe shortage of high-quality Token supply, leaving enormous untapped opportunities in this track.

2. A proven mature business model for reference: As the foundational layer for all AI applications, the Token Factory has a clear business model. Founded less than three years ago, Silicon Based Flow has achieved over 10x revenue growth and rapidly scaled its overseas business, validating the commercial viability of the model. Sellers can follow this path to enter the track.

3. Risk and opportunity insights: The core risks for AI deployment currently stem from unstable underlying Token supply and excessively high costs. Sellers building upper-layer AI applications can partner with established third-party MaaS platforms to avoid underlying supply chain risks, reduce deployment costs via ecosystem resources, and capture the dividends of the AI boom.

Below are key insights for manufacturing factories on digital transformation and business opportunity tapping:

1. Demand reference for AI-enabled production transformation: As AI-enabled production and digital-intelligent transformation advance across all industries, there is massive demand for low-cost, compliant underlying AI capabilities. Silicon Based Flow offers privately deployed Token Factories that meet the strict data compliance requirements of state-owned enterprises and manufacturing firms, helping factories deploy AI at low cost.

2. Commercial opportunities for idle resources: Factories that have built intelligent computing centers or hold idle computing power can leverage Silicon Based Flow's solutions to convert raw computing power into a standardized Token Factory with one click. This solves pain points including difficult adaptation for heterogeneous computing, low utilization rates and poor return on assets, maximizing the commercial value of computing assets.

3. Insights for digital transformation: Factories do not need to invest in developing underlying AI infrastructure from scratch to advance AI transformation. They can leverage the capabilities of mature third-party MaaS platforms to lower technical barriers and capital expenditure, and quickly integrate AI into production, management, sales and other core business links.

Below are key industry insights for AI-related service providers:

1. Industry development trends: The AI industry has now reached a historic inflection point for the explosion of the Token economy. As core infrastructure for the intelligent age, equivalent to utilities in the traditional economy, the Token Factory has enormous market potential. Its business model also carries far more certainty than unproven upper-layer AI applications, making it a high-potential track for AI service providers to enter.

2. Core customer pain points in the industry: The core barrier to large-scale AI deployment across industries is structural computing power shortage, fragmentation and low utilization, which leads to a severe shortage of high-quality Token supply and drives up Token costs, holding back AI adoption. Service providers that solve this pain point will unlock massive market opportunities.

3. Reference for proven deployable solutions: Service providers can follow Silicon Based Flow's path: conduct atomic-level engineering optimization via self-developed inference engines to maximize computing power efficiency, then build a heterogeneous computing power management and scheduling system to unify the scheduling of diverse heterogeneous computing power and improve utilization, and quickly adapt to all mainstream models. This delivers a deployable solution that meets industry demand.

Below are key takeaways for AI-related platform players:

1. Core industry demand for platforms: Massive, cost-effective and stable Token supply demand is about to explode across all industries, and existing supply cannot close the market gap. This creates a major development opportunity for AI infrastructure platforms, and布局 the Token Factory track aligns with broader industry development trends.

2. A reference for clear, replicable platform growth paths: Silicon Based Flow's growth path is replicable: start by developing a large model inference engine to build a solid technical foundation, then launch public cloud MaaS and private MaaS sequentially to cover the needs of different customer segments, then roll out an elastic computing scheduling engine to improve user experience, while partnering with capital and ecosystem players across the industry chain to achieve rapid growth.

3. Risk mitigation guidance: To布局 this track, platforms need to resolve core technical challenges including multi-chip adaptation and heterogeneous computing scheduling in advance. They must also offer both public cloud services and private deployment capabilities to meet the data compliance requirements of different customers, solve pain points including fragmented computing power and low utilization, and mitigate technical and market risks.

Below are the latest industry insights for AI industry researchers:

1. Latest industry developments: The AI industry has reached a historic inflection point and entered the full-scale outbreak phase of the Token economy. As the certain foundational layer of the AI industry, the Token Factory has become a hot track for布局 across the full industry chain. Leading startup Silicon Based Flow has secured co-investment from top industry players across energy, chips, computing power and large models. It has achieved extremely fast commercial growth and is now ranked among the world's leading third-party MaaS platforms.

2. New emerging industry problems: The current development of the AI industry has exposed multiple new problems, including structural computing power shortage, fragmentation, low utilization, severe shortage of high-quality Token supply, and elevated Token costs that constrain large-scale AI deployment. All of these are worthy of in-depth research.

3. Research value of the innovative business model: The Token Factory is an innovative business model born from the division of labor in the AI industry. It packages computing power, algorithms and system capabilities into standardized Tokens for output as core infrastructure of the intelligent age. With a clear business model and validated commercial viability, it opens up a new direction for specialized division of labor in the AI industry and carries high research value.

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.

近日,硅基流动已完成超20亿元B轮融资。本轮融资由携程战投、晶科能源控股、金蝶股份、联通新沃、联通资本旗下联创创新基金、盛奕资本(润泽集团产业资本)、壁仞战投、蔚来资本、商汤战投、巨人网络、国泰君安创新投、纪源资本、华熙睿登、华控基金、国开金融、北工投资、中关村科学城等产业资本、顶级财务机构和国资联合投资,华兴资本担任独家财务顾问。至此,公司已获得能源、芯片、算力基础设施、云服务、大模型及场景应用全产业链头部企业的投资。

过去一年,公司在企业级市场实现爆发式增长:通过Token工厂模式提供高效MaaS(模型即服务),日均Token(词元)调用量达数万亿,服务超1,000万用户和1万家企业客户,营收同比增长超10倍,海外市场单月营收达数百万美元。

标杆客户涵盖能源、金融、交通等核心行业头部央企,公共事业及国家级科研院所,多家电信运营商与地方大型智算中心,云服务巨头、头部大模型企业与互联网大厂,以及数百个头部AI应用与开发工具。

多项第三方数据表明,硅基流动的市场影响力已跻身行业第一梯队,并成为中国最知名的第三方MaaS平台:

国内市场稳居第一梯队。Gartner最新发布的《全球AI云市场指南》显示,硅基流动是入选代表性厂商的少数中国公司之一,同时也是《中国AI基础设施市场指南》中入选代表性厂商的唯一一家创业公司;IDC最新发布的《中国AI软件市场半年度追踪,2025H2》显示,硅基流动是中国公有云MaaS市场份额前四中唯一一家创业公司;Similarweb数据则表明,硅基流动的网站流量与头部大厂规模相当,是同类型创业公司的数十倍。

全球影响力快速崛起。知名Token分发平台OpenRouter数据显示,在70余家Token供应商中,SiliconFlow的日Token消耗量连续数周位居第一,模型平均缓存命中率、输出速度等关键指标长期名列前茅。知名AI开发平台Dify Marketplace数据显示,SiliconFlow的插件安装量超55万,在来自全球的100余家Token供应商中排名前五。

Token工厂,AI时代确定性最高的商业机遇

大模型掀起的是一场生产力革命,其深远程度堪比蒸汽机与电力,将重塑所有产业。这场工业革命造就了两类商业机遇:一类是贴近场景的各类应用,百花齐放但面临高度不确定性;另一类则是所有应用都赖以生存且确定性极高的底座——无论上层应用如何变迁,都离不开底层能力的持续供给。Token工厂正是这个确定性极高的底座。

AI底座的核心技术要素是芯片、模型与系统软件。它们并非孤立存在,而是深度耦合,最终凝练成一种全新的产品形态——Token工厂。它将算力、算法与系统能力封装起来,大规模、低成本、高可靠地输出“智能”的原子单位Token。

作为智能的生产与分发枢纽,Token工厂将成为未来智能社会的底层基础设施,扮演新时代“水电煤”的角色。过去数十年,公路、铁路、电网、通信网络等基建构成了经济腾飞的基石。而未来,国家与企业的竞争力将极大取决于智能的生产和输送能力。

大模型能力每迈上一个台阶,就解锁一大类应用,应用反过来又提升了底层Token消耗量。IDC预测,2026年中国市场的Token消耗量将达到40,000万亿,较2025年再增约20倍。

面对海量的市场需求爆发,大规模、高性价比Token的稳定供应,正成为影响AI技术规模化落地的关键能力。然而,算力的结构性紧缺、碎片化以及利用率低下,导致优质Token供给严重不足;同时,随着大模型解锁更多应用场景,高企的Token成本成为制约企业规模化落地的核心痛点。

硅基流动,Token工厂的最早践行者

硅基流动成立之初,即前瞻性地提出Token工厂模式,致力于提升Token生产效率并通过规模化生产实现技术普惠,现已通过一系列里程碑逐步把最初的目标变成现实:

·2023年8月启动研发大模型推理引擎,已成为世界上少数经过大规模生产环境验证的引擎之一。

·2024年5月,公有云MaaS上线。上架数十款主流开源模型,打通从“裸金属”到标准化Token的链路,并推进支持多元化算力的战略。

·2025年2月,基于华为昇腾的DeepSeek推理服务上线,实现业内首个基于国产芯片的超大规模Token生产服务。硅基流动现已成为支持国产芯片数量最多、效果最好、使用最广的Token生产线。

·2025年9月,私有化MaaS推出。针对拥有自有算力、对数据合规有极高要求的国央企、金融机构,支持在其私有环境中快速搭建专属Token工厂。

·2026年4月,新一代算力调度引擎“弹性GPU”上线,实现多元异构算力的弹性扩缩容,支持客户自助式高效部署模型。

经过三年深耕,硅基流动已在Token生产线各环节全面领跑,并成为具有全球影响力的Token工厂。

世界领先的Token生产线,全面满足产业需求

硅基流动团队深耕AI系统软件方向十余年,以不弃微末、久久为功的工匠精神,潜心技术攻关和产品打磨,构建了世界领先的全自研Token生产线。

“原子级”工程优化,极致释放算力效能:自研推理引擎融合PD分离、KV缓存管理、专家并行、流水并行等领先技术,支持DeepSeek、Qwen、GLM、Kimi等主流模型,可在英伟达、昇腾、沐曦、摩尔线程等多元芯片上稳定提供高吞吐、低延迟的商业级服务。

异构算力纳管与调度,破解供需匹配难题:对异构集群实现统一智能调度与弹性伸缩,支持跨地域资源调度,显著提升算力利用率并实现极致成本管控,满足企业级客户的业务多元化与Token量产需求。硅基流动已与国内外众多算力供应商建立战略合作关系。

超160款模型快速适配,提供全模态选择:将不同模型的适配能力沉淀为可复用模块,实现最新开源模型的快速部署。平台适配模型数量在第三方供应商中大幅领先,覆盖全模态任务,为企业应对复杂场景提供充足选择。

基于自研推理引擎与异构算力纳管调度系统,硅基流动的解决方案可将任何裸算力一键转化为标准化Token工厂,助力算力持有方全面解决异构算力适配复杂、利用率不高、资产回报率低等核心痛点,实现算力资源商业价值的最大化。

产业巨头联合投资,Token供应链深度协同

本轮融资的投资方阵容贯穿AI产业链上下游,这种全产业链投资阵容,不仅带来资金支持,更带来场景、算力、模型和市场的深层生态协同。

晶科能源控股表示:“晶科能源以‘光伏+储能’解决方案助力AI智算中心提供绿色低成本电力;硅基流动凭借跨芯片适配、推理优化等核心能力整合多元算力,从算力侧提升Token产出效率。这次投资是双方在资本与产业层面强强联合,打通‘电力-算力-模型-平台-应用’全链路,打造绿色Token工厂,助力AI算电协同基建可持续发展与规模化落地。”

金蝶股份表示:“硅基流动在AI基础设施领域的全栈闭环能力,与我们推进AI优先、构建企业AI生态布局的战略方向高度契合。金蝶深耕企业管理云服务三十余年,积累了丰富的行业Know-How与海量高价值业务场景。我们期待与硅基流动围绕AI基础设施、模型服务、企业级应用场景及生态共建展开深度协同,助力企业以更低成本、更高效率获取和应用AI能力,推动企业AI从技术创新走向规模化落地与商业价值创造。”

联通创投旗下出资平台表示:“联通创投以资本手段服务集团主责主业,助力算网融合创新与AI全融通计划,助力AI基建、AI科技、AI应用的融合创新与生态建设。本次投资进一步完善AI全链条生态、构建‘连接+算力+AI服务’全栈能力体系,助力跨地域异构算力协同,促进国产化、低成本、一站式算网融合解决方案形成,面向政务、工业、开发者输出普惠AI算力,助力培育数字经济新质生产力。”

润泽集团表示:“硅基流动自研的推理引擎和异构算力纳管调度,与润泽集团当前布局的液冷智算集群具有较强产业协同效应。我们看好其技术壁垒与商业化落地能力,打通了模型、芯片、算力、工具链与终端应用生态,助力构建多方共赢的跨生态协同体系。双方将携手打造满足多行业多场景客户需求的一站式MaaS算力服务解决方案。”

壁仞科技投资负责人表示:“硅基流动是率先在多家国产算力芯片上实现‘推理部署-算力调度-应用落地’闭环的AI基础设施企业,我们高度认可其深厚技术积累和工程化落地实力,期待在芯片适配、推理加速优化及算力集群规模化部署落地等层面开展深度合作,共同打造高性能Token工厂,为更多企业和开发者提供更可控、更易用、更高效的国产AI算力解决方案。”

国泰君安创新投投资负责人表示:“我们长期看好AI Infra赛道的发展潜力,充分认可公司的技术实力与发展前景,并看好开源生态与国产芯片背景下的中间调度层应用价值。硅基流动在核心技术领域构建了突出壁垒,商业化增长迅速、路径清晰,其发展布局与推动AI产业落地的战略高度契合。”

纪源资本表示:“我们相信,能够真正打通算力供给与模型应用、让普惠算力触手可及的公司,将拥有巨大的长期价值。袁进辉博士在分布式系统与异构算力领域有着深厚的技术积累与前瞻判断,其带领的硅基流动团队技术功底扎实、工程能力出众,并在国产算力与大模型落地的产业生态中占据了独特而关键的位置。他们正在打造国产算力时代的Token工厂,期待公司持续拓展算力效率的边界,让高效、普惠的智能算力真正服务于千行百业。”

中关村科学城表示:“我们选择投资硅基流动,看中其顶尖技术团队、差异化算力底座路线与国产算力适配的长期价值。依托科学城全链条创新生态,我们将全方位赋能硅基流动在政企、制造、文娱等多元场景落地。”

华兴资本投资银行事业部董事高鹏飞表示:“非常荣幸能够陪伴硅基流动从天使轮一路走来,见证一个又一个重要里程碑。2026年,AI产业链迎来历史性拐点,Token经济全面爆发。不到三年时间,硅基流动从打造最快的推理引擎,成长为国内最大的第三方MaaS平台;从服务千万级开发者,到覆盖超万家企业客户;从国内头部Token工厂,到登顶全球OpenRouter排名榜首。我们深度体会过这支团队的战斗力与凝聚力,也坚信硅基流动将在未来不断创造新的奇迹。本轮融资获得众多产业链上下游重量级战略方的加持,也必将进一步助力公司实现‘加速AGI普惠人类’的美好愿景。”

结语

Token经济的爆发,是AI时代确定性最大的商业机遇。硅基流动凭借前瞻性投入和战略定力,专注于Token生产核心技术研发,产品和服务品质获得了市场广泛认可。在本轮全产业链上下游巨头资本的联合加持下,硅基流动将向着“全球最好的Token供应平台”加速迈进,为AI技术普惠人类作出自己的贡献。

注:文/龚作仁,文章来源:Laborer,本文为作者独立观点,不代表亿邦动力立场。

文章来源:Laborer

广告
微信
朋友圈

这么好看,分享一下?

朋友圈 分享

APP内打开

+1
+1
微信好友 朋友圈 新浪微博 QQ空间
关闭
收藏成功
发送
/140 0