Moonshot is in early discussions with Microsoft, Amazon and Alphabet’s Google about hosting its Kimi K3 model on Azure, Amazon Web Services and Google Cloud, according to three people familiar with the talks cited by Reuters.
The Beijing-based AI company is seeking as much as 30% of revenue generated from K3-related services on the three cloud platforms, the sources said. The proposed terms are similar to arrangements Moonshot has discussed with other large customers using the model.
For cloud partners, the potential agreements could create another route for delivering lower-cost frontier AI to enterprise customers, although US policy and data-governance concerns could complicate those opportunities.
No agreement has been reached, and the talks could still fall apart. Any completed agreement could become the first major revenue-sharing pact between a Chinese AI developer and a major US cloud provider, Reuters reported.
The negotiations also have some practical hurdles. The companies have yet to settle how revenue would be divided, how Moonshot would access data and how token usage would be audited. Those details matter because token consumption is central to usage-based AI billing.
Why K3 needs the cloud
Kimi K3 is an open-weight model with 2.8 trillion parameters, giving customers the ability to download and modify its weights. But that openness does not make the model cheap to operate.
Although customers can download and modify K3, operating a 2.8-trillion-parameter model at enterprise scale can require substantial computing capacity, making major cloud platforms an important route to wider adoption. Moonshot has already signed revenue-sharing agreements with smaller cloud providers, while Chinese IT services company Chinasoft International disclosed a similar deal in July without revealing its financial terms.
K3 has also attracted attention in independent evaluations. Arena.ai ranked it first for web interface-building capabilities, while Artificial Analysis found its performance comparable with OpenAI’s GPT-5.5 and Anthropic’s Claude Opus 4.8 on demanding, multi-step tasks, according to Reuters.
A commercial opportunity with political risk
The talks come at an awkward time for Moonshot. US Treasury Secretary Scott Bessent said last month that he might add the company to a trade blacklist. US officials have also accused Moonshot of using outputs from Anthropic’s Fable model to help develop K3 through distillation and of illegally acquiring Nvidia chips, Reuters reported.
Moonshot has rejected the distillation allegations, telling China’s National Business Daily that K3’s gains came from original changes to its underlying architecture.
That leaves the cloud companies facing a difficult calculation: K3 could give their enterprise customers another capable and potentially lower-cost model, but a partnership with Moonshot could attract additional scrutiny from Washington.
What could happen next
For Moonshot, access to Azure, AWS and Google Cloud would be more than another distribution deal. It could turn K3 from an open model that customers can download into a service that is far easier for businesses to deploy.
The proposed 30% revenue share is also significant. It would give Moonshot a sizable portion of the economics while leaving cloud providers responsible for the infrastructure and customer distribution. But the biggest uncertainty is not the percentage. Data access, usage verification and U.S. policy could ultimately determine whether the agreements happen at all.
Moonshot raised more than $2 billion in May and is preparing for a potential Hong Kong listing. A successful US cloud rollout could strengthen K3’s position as a global commercial platform while giving channel partners another model to deploy and support. However, partners would still need to weigh data governance, contract continuity and the risk that future US restrictions could disrupt customer deployments.
Read more: AWS, Microsoft and Google Cloud are converging around AI-led growth, creating new opportunities for partners while raising expectations around security, governance and scale.