Beijing-based artificial intelligence developer Moonshot AI is negotiating revenue-sharing partnerships with Microsoft, Amazon, and Alphabet's Google to host its flagship open-weight model, Kimi K3, across major cloud platforms. According to reporting from Reuters, the startup is seeking up to a 30 percent share of revenue generated from Kimi K3 inference services hosted on Microsoft Azure, Amazon Web Services (AWS), and Google Cloud.
Moonshot released Kimi K3 in July 2026 as a 2.8-trillion parameter open-weight model featuring a 1-million-token context window. While the model weights are publicly downloadable for research and self-hosting, the sheer scale of a multi-trillion parameter architecture makes private infrastructure hosting cost-prohibitive for most enterprise organizations, establishing hyperscaler cloud catalogs as the principal distribution channel for enterprise adoption.
Open-Weight Licensing and Commercial Monetization
Moonshot's commercial strategy relies on provisions embedded in Kimi K3's licensing terms. Under the model's license agreement, any commercial provider offering Kimi K3 as a managed service and generating over $20 million in annual gross revenue must establish a formal commercial agreement with Moonshot.
The proposed 30 percent revenue split aligns with the licensing tiers Moonshot has presented to enterprise software vendors and regional infrastructure operators. Chinese IT services provider Chinasoft International previously disclosed a commercial revenue-sharing partnership with Moonshot in a regulatory filing, though the precise split was not publicly detailed.
Key points currently under discussion between Moonshot and US cloud providers include:
- Revenue distribution percentages across managed API calls and dedicated instance hosting.
- Verification mechanisms and data audit protocols for tracking token consumption across third-party cloud billing systems.
- Compliance and governance boundaries regarding enterprise data security and regional sovereign hosting requirements.
Hyperscaler Economics for Trillion-Parameter Models
The negotiations reflect a shifting dynamic between foundational model developers and cloud infrastructure giants. As frontier open-weight models reach trillions of parameters, model creators are using tiered commercial licenses to capture direct monetization from the massive inference volumes generated on public cloud platforms.
Moonshot, which has secured backing from Alibaba and Tencent and is preparing for a potential Hong Kong initial public offering, has also explored similar revenue-sharing arrangements with regional cloud operators. The discussions with US cloud providers remain in early stages, with formal commercial terms still subject to negotiation.
