Moonshot AI is scaling both capital and enterprise distribution for its Kimi K3 model.

Moonshot AI is scaling both capital and enterprise distribution for its Kimi K3 model. On July 30, 2026, Technode reported that Kimi K3 is now available through Fireworks AI on Microsoft Foundry, giving Azure customers a managed route to run the model through OpenAI-compatible Chat Completions and Responses APIs. Fireworks supplies the inference engine and dedicated capacity, while Moonshot remains the model provider. On the same day, Bloomberg reported that Moonshot closed about $3.5 billion in funding at a $35 billion valuation, roughly double the initial target, and that annual recurring revenue rose from $100 million in March to $300 million in June. Pandaily coverage indicates Kimi K3 captured 41% of global open-source downloads within two days of launch and drove 4x paid user growth overseas. A Hacker News post demonstrates running full Kimi K3 with 29GB of RAM, pointing to aggressive inference efficiency. A Bloomberg video notes Kimi was built with Nvidia compute.

What the sources said

• "Microsoft Foundry provides the Azure-native environment for procurement, identity, billing and governance, while Fireworks supplies the inference engine and dedicated capacity." — Technode
• "The company’s annual recurring revenue reportedly rose from $100 million in March to $300 million in June, adding a business-growth metric to the valuation story." — Technode via Bloomberg
• "Kimi K3 Open-Source Model Goes Global With 41% of World Open-Source Downloads and 4x Paid User Growth Overseas." — Pandaily headline

This combination signals higher demand for Nvidia-based inference capacity on Azure as enterprises adopt Kimi K3 through a managed route, while the funding round provides Moonshot with capital to secure compute and expand its model roadmap.