广告
加载中

月之暗面接洽美三大云厂商 最高拿三成收入分成

亿邦AI 2026-08-27 14:06
亿邦AI 2026/08/27 14:06

邦小白快读

EN
全文速览

本文主要披露了中国AI初创企业月之暗面与美国三大云厂商合作洽谈的核心信息,干货整理如下:

1. 核心合作内容:月之暗面正和微软、亚马逊、谷歌三大美国云厂商开展合作洽谈,若协议落地,三家将分别在自有云平台托管月之暗面的Kimi K3大模型;该模型是月之暗面预计2026年7月发布的旗舰基座模型,参数规模达2.8万亿,曾拿下全球代码竞技场评测榜单第一名。

2. 合作进展与诉求:本次洽谈仍处于早期阶段,最终合作尚未敲定,待确认细节包括收入拆分比例、数据访问规则、token使用量统计方式等;月之暗面提出最高可拿到K3相关服务30%的收入分成,此前从未有中国AI企业和美国主流云服务商达成过同类合作。

3. 现有风险情况:目前美国相关人士放出针对月之暗面的贸易禁令威胁,美国相关部门指控该公司窃取同行模型技术,月之暗面已经否认所有指控,三大云厂商暂未做出回应。

本文透露出中国AI大模型品牌出海合作的新动向、风险与可参考经验,干货整理如下:

1. 出海合作新模式:本次是中国AI大模型品牌首次尝试和美国三大顶级云厂商洽谈模型托管+收入分成的合作,月之暗面开出最高三成收入分成的条件,打破了此前没有同类合作的局面,给出海AI品牌提供了新的合作路径参考。

2. 出海需警惕的风险:中国AI品牌出海目前面临明确的地缘政治风险与知识产权风险,美国已经出现针对月之暗面的贸易禁令威胁和知识产权指控,这类风险会直接影响合作落地,所有计划出海的AI品牌都需要提前预警、做好应对预案。

3. 产品研发方向参考:月之暗面能拿到对接三大云厂商的机会,核心依托的是Kimi K3大模型的技术实力,该模型参数规模领先、评测成绩全球第一,说明技术硬实力是打开海外市场的核心竞争力,品牌研发需要优先聚焦技术提升。

针对布局AI业务出海的卖家,本文透露出新的行业机会与需要警惕的风险,干货整理如下:

1. 新的出海机会:中国AI大模型首次尝试切入美国主流云服务生态,本次洽谈如果最终落地,会打通一条全新的中国AI产品出海路径,给国内中小AI相关卖家提供可复制的合作模板,后续会有更多切入海外市场的机会。

2. 需要提前规避的风险:目前中国AI企业出海美国面临明确的政策风险与知识产权风险,已经出现贸易禁令威胁和知识产权指控,这类风险会直接导致合作停滞,卖家出海前需要做好合规审查与风险应对预案,避免踩坑。

3. 可参考的新商业模式:本次合作探索的“云平台托管模型+收入分成”模式,属于轻资产出海模式,卖家不需要自行搭建海外算力基础设施,只需要输出模型技术就能获得收益,对中小AI卖家来说非常友好,具备参考价值。

对于对接AI产业的国内工厂,本文透露出新的商业机会与数字化转型启示,干货整理如下:

1. 产品生产端的新需求:当前AI大模型的参数规模不断提升,本次月之暗面推出的旗舰模型参数已经达到2.8万亿,大模型对算力硬件的性能、产能要求都在不断提升,相关AI硬件生产工厂可以提前调整产能布局,匹配市场需求。

2. 新的商业机会:中国AI大模型的出海进程正在加速,本次和美国三大云厂商的合作如果落地,会带动整个AI产业链的出海配套需求,给国内做AI硬件生产、配套零部件供应的工厂带来更多海外订单,拓展新的增长空间。

3. 数字化转型的启示:AI大模型商业化选择和专业云平台合作降低自身投入,这种模式对工厂数字化转型同样有参考价值,工厂推进数字化不需要完全自建技术团队,可以依托专业平台的能力降低转型成本,提升转型效率。

对于服务AI产业的服务商,本文透露出行业新趋势与待解决的痛点,干货整理如下:

1. 行业发展新趋势:当前中国头部AI大模型厂商已经开启全球化商业化探索,和美国顶级云厂商合作分成是全新的商业化路径,AI大模型的全球化分工体系正在逐步形成,给国内AI服务商带来了新的市场空间,可拓展出海相关服务业务。

2. 行业现存客户痛点:本次合作中还有多个核心问题待解决,包括收入拆分规则制定、跨境数据访问规则适配、token使用量统计标准统一等,这些都是AI大模型出海合作中普遍存在的痛点,服务商可以针对这些痛点开发对应解决方案,抢占新市场。

3. 技术升级的方向:当前大模型参数规模不断提升,技术要求越来越高,对算力服务、合规服务、技术服务的要求也在提升,服务商需要不断迭代自身技术能力,才能匹配行业最新的需求,获得更多客户。

对于云服务等相关平台商,本文透露出大模型厂商的合作需求与行业潜在风险机会,干货整理如下:

1. 大模型厂商的最新合作需求:当前AI大模型厂商更倾向于通过入驻云平台托管模型的方式实现商业化,更接受按收入分成的结算模式,本次月之暗面提出最高三成分成的要求,反映了大模型厂商对合作收益的核心诉求,平台可以调整自身合作政策,吸引更多优质大模型品牌入驻。

2. 需要规避的潜在风险:平台引入海外大模型会面临一定的政策合规风险,本次合作中美国已经出现贸易禁令威胁和知识产权指控,平台在招商引入海外大模型的时候,需要提前做好合规评估,规避政策风险,避免给平台带来损失。

3. 新的招商机会:目前中国不少大模型已经达到全球顶级技术水平,具备很强的竞争力,引入优质中国大模型可以丰富平台的产品矩阵,提升平台的整体竞争力,是平台招商新的增长点。

对于AI产业研究者,本文披露了全球AI产业合作的新动向与新问题,研究价值较高,干货整理如下:

1. 产业发展新动向:本次是首次出现中国AI初创企业和美国三大顶级云厂商洽谈大模型托管分成合作,打破了此前没有同类合作的局面,说明中国AI大模型的技术实力已经获得国际行业的关注,中国AI企业的全球化商业化探索进入全新阶段,是值得研究的产业新趋势。

2. 产业出现的新问题:当前中美AI产业合作受到地缘政治因素影响,贸易禁令威胁、知识产权指控都成为合作落地的阻碍,同时全球范围内还没有形成成熟的大模型跨地域托管合作规则,分成比例、数据规则、用量统计都没有通用标准,这些都是产业发展中出现的新问题,具备较高的研究价值。

3. 商业模式研究新样本:本次探索的云平台托管+收入分成模式,是大模型商业化的全新路径,给研究AI大模型的商业化模式提供了新的研究样本,对总结大模型商业化规律有重要参考意义。

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 discloses core details regarding ongoing cooperation talks between Chinese AI startup Moonshot AI and three major U.S. cloud providers, with key takeaways as follows:

1. Core cooperation scope: Moonshot AI is in discussions with Microsoft, Amazon and Google to host its Kimi K3 large language model on each provider’s respective cloud platform. K3 is Moonshot’s flagship base model scheduled for release in July 2026, with 2.8 trillion parameters, and it currently ranks first on the global code arena benchmark leaderboard.

2. Progress and demands: Talks remain in the early stage, with no final agreement reached. Unresolved details include revenue split ratios, data access rules and token usage measurement methods. Moonshot AI is asking for up to 30% of revenue generated from Kimi K3-related services; no Chinese AI company has ever reached a similar cooperation agreement with mainstream U.S. cloud service providers before.

3. Current risks: U.S. stakeholders have threatened to impose trade bans on Moonshot AI, and U.S. regulators have accused the company of stealing model technology from industry peers. Moonshot AI has denied all allegations, and the three cloud providers have not yet issued any public response.

This article outlines new trends, risks and actionable insights for Chinese AI large model brands looking to expand overseas, with key takeaways as follows:

1. A new outbound cooperation model: This marks the first time a Chinese AI large model brand has negotiated a model-hosting-and-revenue-sharing partnership with the three top U.S. cloud providers. With Moonshot AI requesting up to a 30% revenue split, this breakthrough breaks the previous barrier of no comparable cooperation, offering a new viable cooperation path for other outbound AI brands.

2. Risks to watch for: Chinese AI brands expanding to the U.S. market face tangible geopolitical and intellectual property risks. The trade ban threat and IP allegations targeting Moonshot AI directly threaten the success of cooperation, requiring all AI brands planning outbound expansion to prepare risk mitigation plans in advance.

3. R&D guidance: Moonshot AI secured talks with the three major cloud providers primarily thanks to the technical strength of the Kimi K3 model, which boasts industry-leading parameter scale and the top global benchmark ranking. This confirms that strong technical capability is the core competitive advantage for entering overseas markets, so brands should prioritize technical improvement in R&D.

For sellers looking to expand AI-related business overseas, this article outlines new industry opportunities and key risks to avoid, with key takeaways as follows:

1. New outbound opportunities: This is the first time a Chinese large AI model has attempted to integrate into the mainstream U.S. cloud service ecosystem. If the cooperation is finalized, it will open up an entirely new outbound path for Chinese AI products, creating a replicable cooperation template for small and medium-sized Chinese AI sellers and unlocking more future access to overseas markets.

2. Pre-emptive risk mitigation: Chinese AI companies expanding to the U.S. face clear policy and intellectual property risks, as demonstrated by the ongoing trade ban threat and IP allegations against Moonshot AI, which could derail cooperation entirely. Sellers must complete compliance reviews and prepare risk response plans before entering overseas markets to avoid costly missteps.

3. A replicable new business model: The "cloud-hosted model + revenue sharing" model explored in this cooperation is an asset-light outbound approach. Sellers do not need to build their own overseas computing infrastructure, and can generate revenue simply by licensing their model technology, making it extremely accessible and valuable for small and medium-sized AI sellers.

For Chinese factories serving the AI industry, this article outlines new business opportunities and insights for digital transformation, with key takeaways as follows:

1. New demand on the production side: The parameter scale of large AI models continues to grow, with Moonshot AI’s flagship K3 model already reaching 2.8 trillion parameters. Large models are driving rising requirements for both performance and production capacity of AI computing hardware, so relevant AI hardware manufacturers can adjust their capacity layout in advance to align with market demand.

2. New business opportunities: The outbound expansion of Chinese large AI models is accelerating. If the cooperation with the three major U.S. cloud providers is finalized, it will drive increased demand for supporting services across the entire AI industry chain, bringing more overseas orders and new growth opportunities for Chinese factories producing AI hardware and supplying supporting components.

3. Insights for digital transformation: Large model developers choose to partner with professional cloud platforms to reduce capital investment, a model that holds lessons for factory digital transformation. Factories do not need to build entirely in-house technical teams to advance digitalization; they can leverage the capabilities of professional platforms to cut transformation costs and improve efficiency.

For service providers serving the AI industry, this article outlines new industry trends and unaddressed pain points, with key takeaways as follows:

1. New industry growth trends: Leading Chinese large AI model developers have begun exploring global commercialization, and revenue-sharing cooperation with top U.S. cloud providers represents an entirely new commercialization path. A global division of labor for large AI models is gradually taking shape, creating new market space for Chinese AI service providers to expand outbound-related services.

2. Existing customer pain points: Multiple core issues remain unresolved in the current cooperation talks, including revenue split framework, cross-border data access rule adaptation, and unified standards for token usage measurement. These are common pain points for cross-border large model cooperation, and service providers can develop targeted solutions for these issues to capture new market share.

3. Priorities for technical upgrading: As large model parameter scales continue to grow and technical requirements rise, demand for higher-quality computing, compliance and technical services is also increasing. Service providers must continuously iterate their technical capabilities to align with the latest industry requirements and win more clients.

For cloud service and other relevant platform providers, this article outlines the cooperation demands of large model developers and the potential industry risks and opportunities, with key takeaways as follows:

1. Latest cooperation demands from large model developers: Large AI model developers now increasingly prefer to commercialize their products by hosting models on cloud platforms, and are increasingly accepting revenue-sharing settlement models. Moonshot AI’s request for up to a 30% revenue split reflects the core profit demand of large model developers in cooperation, so platforms can adjust their cooperation policies to attract more high-quality large model brands.

2. Potential risks to mitigate: Platforms introducing foreign large models face certain policy and compliance risks. Given the trade ban threat and IP allegations in this case, platforms must complete compliance assessments in advance when onboarding overseas large models to mitigate policy risk and avoid losses for the platform.

3. New growth opportunities for business development: Many Chinese large models have already reached top-tier global technical standards and boast strong competitiveness. Onboarding high-quality Chinese large models can enrich a platform’s product matrix and improve its overall competitiveness, making it a new growth driver for platform business development.

For AI industry researchers, this article discloses new trends and emerging issues in global AI industry cooperation that carry high research value, with key takeaways as follows:

1. New industry development trends: This is the first time a Chinese AI startup has held large model hosting and revenue sharing talks with the top three U.S. cloud providers, breaking the previous absence of such cooperation. This signals that the technical capability of Chinese large AI models has gained attention from the international industry, and Chinese AI companies have entered a new stage of global commercial exploration, making this a noteworthy new industry trend for research.

2. New emerging industry issues: Sino-U.S. AI industry cooperation is heavily impacted by geopolitical factors, with trade ban threats and IP allegations creating barriers to finalizing deals. At the same time, no mature standardized framework for cross-border large model hosting cooperation exists globally, with no universal standards for split ratios, data rules or usage measurement. These new emerging issues in industry development carry high research value.

3. A new research sample for business models: The cloud hosting + revenue sharing model explored in this cooperation is an entirely new commercialization path for large models, providing a fresh research sample for academic and industry analysis of large model commercialization, and offering important reference value for summarizing patterns in large model commercialization.

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.

中国AI初创企业月之暗面正与微软、亚马逊、Alphabet旗下谷歌开展合作洽谈,若协议落地,三家美国云服务商将分别在旗下Azure、AWS、Google Cloud平台托管月之暗面的Kimi K3大模型。Kimi K3是月之暗面2026年7月发布的旗舰基座模型,参数规模达2.8万亿,曾登顶全球代码竞技场评测榜单首位。

三位熟悉洽谈进程的人士透露,月之暗面要求最高获得K3相关服务30%的收入分成。此前还没有中国AI企业与美国主流云服务商达成过同类合作。

目前相关洽谈仍处于早期阶段,最终合作尚未敲定,待确认的事项包括收入拆分比例、数据访问规则、token使用量统计方式等内容。

近期美国财政部长Scott Bessent曾放出针对月之暗面的贸易禁令威胁,美国相关部门指控该公司窃取Anthropic旗下Fable模型,月之暗面否认所有相关指控。微软、谷歌、AWS三方暂未就相关洽谈作出回应。

本文首发于 亿邦动力 官方网站

文章来源:亿邦动力

广告
微信
朋友圈

FAQ回顾

月之暗面与美国三大云厂商的洽谈内容是什么?

中国AI初创企业月之暗面正与微软、亚马逊、Alphabet旗下谷歌开展合作洽谈,若协议落地,三家美国云服务商将分别在旗下Azure、AWS、Google Cloud平台托管月之暗面的Kimi K3大模型,月之暗面要求最高获得K3相关服务30%的收入分成。

Kimi K3大模型的参数和性能表现怎么样?

Kimi K3是月之暗面2026年7月发布的旗舰基座模型,参数规模达2.8万亿,曾登顶全球代码竞技场评测榜单首位,是月之暗面的核心大模型产品。

此前有中国AI企业和美国主流云服务商达成同类合作吗?

此前还没有中国AI企业与美国主流云服务商达成过同类大模型托管合作,目前本次洽谈仍处于早期阶段,最终合作尚未敲定,待确认的事项包括收入拆分比例、数据访问规则、token使用量统计方式等内容。

美国相关方对月之暗面有哪些相关指控?

近期美国财政部长Scott Bessent曾放出针对月之暗面的贸易禁令威胁,美国相关部门指控该公司窃取Anthropic旗下Fable模型,月之暗面已否认所有相关指控,微软、谷歌、AWS三方暂未就相关洽谈作出回应。

这么好看,分享一下?

朋友圈 分享

APP内打开

赞 +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