Moonshot AI Seeks 30% Revenue Sharing with Microsoft, Amazon, Google for Kimi K3 Model

6 Sources

Share

Beijing-based Moonshot AI is in early-stage negotiations with Microsoft, Amazon, and Google over revenue-sharing agreements that would allow US cloud giants to host its Kimi K3 AI model. The startup seeks up to 30% of revenue from Azure, AWS, and Google Cloud services, marking a potential first for Chinese AI firms despite geopolitical tensions and Treasury Secretary warnings.

Moonshot AI Pushes for Historic Revenue-Sharing Deal with US Cloud Giants

Moonshot AI is negotiating revenue-sharing agreements with Microsoft, Amazon, and Google that would enable US cloud providers to host its Kimi K3 AI model on Azure, AWS, and Google Cloud

1

2

. The Beijing-based startup seeks up to 30% of revenue generated from K3-related services, a figure that matches terms the company has outlined for major customers using the open-weight model

1

. If finalized, this would mark the first major revenue-sharing pact between a Chinese AI firm and US cloud platforms, fundamentally reshaping how Chinese AI models reach Western enterprise customers

1

.

Why This Deal Matters for Enterprise AI Adoption

The talks with Microsoft, Amazon, and Google highlight how Chinese AI models are gaining traction in the US market despite national security concerns and export bans on AI chips to China

1

4

. For major AI model providers like Moonshot AI, cloud platforms represent the primary route to enterprise adoption. Although Kimi K3 is an open-weight model that can be downloaded and modified, analysts indicate few customers will run a system with 2.8 trillion parameters on their own infrastructure due to massive computing costs

1

4

. The open-weight model changes who cloud providers compete with rather than eliminating the need for hosting entirely

2

.

Kimi K3 Performance Drives Commercial Interest

Kimi K3 has demonstrated strong results in third-party benchmarks, with Arena.ai ranking it first for web interface-building capabilities

1

4

. Artificial Analysis reports the AI model delivers performance comparable to OpenAI's GPT-5.5 and Anthropic's Claude Opus 4.8, particularly on tests measuring complex, multi-step tasks

1

4

. The model's competitive performance against frontier models from OpenAI and Anthropic has rattled markets and accelerated interest from hyperscalers seeking capable, cost-effective alternatives for their customers

2

.

Unresolved Issues Complicate Negotiations

The discussions remain at an early stage with no certainty they will result in agreements

1

3

. Unresolved issues in Moonshot AI's negotiations with US cloud companies include how revenue might be split, data access provisions, and auditing token usage

1

4

. Token usage auditing is particularly sensitive since tokens are units of text processed by AI models, and measuring their use is central to calculating revenue under usage-based billing

1

4

. Data access raises immediate questions about what a Chinese company can see of prompts and outputs flowing through Azure, AWS, or Google Cloud, concerns enterprise customers will demand answers to before routing sensitive information

2

.

Geopolitical Tensions Cast Shadow Over Potential Deals

The negotiations with Microsoft, Amazon, and Google occur despite critical comments from senior US officials about Moonshot AI

1

4

. US Treasury Secretary Scott Bessent indicated last month he might add Moonshot AI to a trade blacklist, while US officials have accused the Beijing-based company of stealing from Anthropic's Fable model to create Kimi K3 and illegally acquiring Nvidia chips

1

5

. Moonshot AI has rejected suggestions that Kimi K3's performance resulted from distillation, telling China's National Business Daily its performance gains came from original changes to underlying architecture

1

5

. Three US cloud providers signing revenue-sharing agreements with a company the Treasury Secretary wants blacklisted presents a complex political calculation

2

.

Strategic Implications for AI Commercialization

Source: The Next Web

Source: The Next Web

Founded in 2023 by Carnegie Mellon-trained AI researcher Yang Zhilin, Moonshot AI is backed by Chinese tech giants including Alibaba

1

4

. The startup raised more than $2 billion in May and is preparing for a potential Hong Kong IPO at a $30 billion valuation

1

2

. International distribution through Azure, AWS, and Google Cloud would materially transform the revenue story Moonshot AI presents to investors, especially given its valuation jumped sevenfold in six months from roughly $20 billion

2

. Moonshot AI has already signed similar agreements with smaller cloud platforms, including a July deal with Chinese IT services provider Chinasoft International, though revenue split details were not disclosed

1

4

. Alibaba is also pursuing revenue-sharing agreements with major users of its new open-source AI model, suggesting this approach represents a broader trend in Chinese AI commercialization

1

5

. Whether these talks conclude may depend less on commercial terms than on whether Washington decides such arrangements are permissible, making this a test case for how Chinese AI models access Western markets amid intensifying geopolitical tensions

2

.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved