Moonshot AI is moving fast. The Beijing-based startup behind the Kimi K3 large language model is preparing for an initial public offering on the Hong Kong Stock Exchange, targeting a listing within six months, according to Juro Osawa of The Information. The company has already submitted a listing proposal to investors, signaling this is more than an aspiration.
The timing is deliberate. Kimi K3 launched on July 16, 2026, and the response was immediate enough to temporarily break the company’s own onboarding process. Moonshot AI had to pause new user registrations because demand outpaced what its infrastructure could absorb.
A model with serious numbers behind it
Kimi K3 is not a modest release. The model carries 2.8 trillion parameters, making it the largest open-weight model currently available, and supports a 1-million-token context window. For reference, a 1-million-token window can process roughly 750,000 words in a single pass, about the length of seven average novels at once.
Open weights for K3 were released on July 27, 2026, eleven days after the initial launch. That decision to go open dramatically widened the model’s reach, letting developers integrate it without gating through Moonshot AI’s API directly.
The commercial impact has been sharp. API demand surged following K3’s debut, pushing Moonshot AI’s annual recurring revenue toward $300 million. That figure represents a company that was founded in 2023, making the trajectory genuinely unusual even by the standards of the current AI gold rush.
The mixture-of-experts architecture behind K3 is worth understanding briefly. Rather than activating all 2.8 trillion parameters for every request, the model routes each query through a relevant subset. Think of it like a hospital with hundreds of specialists: instead of consulting every doctor on staff for a sprained ankle, the system sends you directly to orthopedics. The result is high capability without proportionally higher compute costs per query.
A $30B valuation and a funding round closing now
Moonshot AI is closing a funding round at a $30 billion valuation. That number is a significant step up from the $20 billion valuation attached to an earlier round in 2026, when the company raised around $2 billion.
The investor roster reads like a who’s-who of Chinese tech capital. Alibaba, Tencent, and Meituan’s venture arm have all backed the company, giving Moonshot AI both financial runway and distribution leverage. Alibaba and Tencent each operate cloud infrastructure at scale, which matters enormously for a company whose primary product is an API-driven language model with surging demand.
The Hong Kong listing target sits within six months of mid-to-late July 2026 reports, which would put a potential IPO in early 2027. Moonshot AI was founded in 2023, which means the company is attempting a public listing roughly three years after inception.
What the IPO push means for the broader AI market
An IPO from Moonshot AI would be one of the first major public listings from a post-2023 AI-native Chinese startup. The $300 million ARR figure is the critical anchor here. Public market investors have become skeptical of AI companies whose revenue stories rely heavily on future projections rather than current contracts. A company approaching that ARR level with clear API demand driving it has a more defensible story to tell in a roadshow.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

1 hour ago
28









English (US) ·