From 300 million to 35 billion, Moonshot AI enters the global top-tier AI arena with open-source models

Original Author: Sleepy

Kimi has just open-sourced the K3 model weights, and Moonshot AI's latest funding round has also been finalized.

According to an exclusive report from a Star Market Daily reporter, Moonshot AI has completed its Series F financing, raising over $3.5 billion, bringing its post-money valuation to $35 billion.

This round of financing was not initially planned to raise this much money. The report stated that because investor subscriptions exceeded the original target by more than three times, Moonshot AI closed the Series F early. The planned Series G round, originally scheduled to start in August, has also been moved up.

The Series G round will be the Pre-IPO round before Moonshot AI's listing, with a pre-money valuation already rising to $50 billion.

Just a week ago, the market version was that Moonshot AI would close its final round of private financing in August after completing a round with a pre-money valuation of approximately $31.5 billion. Now, the Series F has finally raised over $3.5 billion, the post-money valuation has reached $35 billion, and the next round is no longer waiting until August. Both capital and valuation have arrived faster than planned.

Since its establishment in 2023, Moonshot AI has taken just over three years to push its valuation from $300 million to $35 billion. The ongoing Pre-IPO round has already set the next price tag at $50 billion.

Over a Dozen Rounds in Three Years, Valuation Soars from $300 Million to $35 Billion

Founded in April 2023, Moonshot AI was established by Yang Zhilin, Zhang Yutao, Zhou Xinyu, and Wu Yuxin.

About two months after its founding, the company completed an Angel round of over $200 million, with a post-money valuation of around $300 million. For a startup AI model company yet to officially launch a product, this was a remarkably large early-stage funding.

What truly placed Moonshot AI at the center of the capital game table was its Series A+ round completed in February 2024.

This round raised over $1 billion, led by Alibaba, with participation from institutions and industrial capital like Sequoia China, Xiaohongshu (RedNote), and Meituan, pushing Moonshot AI's valuation to approximately $2.5 billion. A subsequent Series B round six months later raised over $300 million, increasing the company's valuation to $3.3 billion.

However, after the 2024 fundraising, Moonshot AI's financing pace slowed down for a while.

During that year, the Chinese large model market underwent significant changes. Tech giants like ByteDance and Alibaba continuously pushed down model prices. DeepSeek rose rapidly with its open-source approach and cost efficiency. Competition in general chat products gradually shifted from user growth to model capability, inference costs, and commercial revenue. Kimi initially stood out with its long-context feature, but relying on a single product differentiator struggled to support higher valuations.

It wasn't until the end of 2025 that Moonshot AI completed a $500 million Series C round, reaching a post-money valuation of $4.3 billion. After this, the company's financing pace accelerated markedly.

In the first two months of 2026, Moonshot AI completed multiple funding rounds in succession, with its valuation rising from $4.3 billion to $10 billion, and further to $18 billion.

In May of this year, the company completed a Series D round of approximately $2 billion, achieving a post-money valuation of $20 billion. Participants were no longer limited to internet companies and market-oriented investment institutions; China Mobile, Guozhi Investment, CPE Source, and several state-backed funds also joined the shareholder list.

Shortly after the Series D concluded, a new round of financing was initiated in June. The market's pre-money valuation at the time had already reached $31.5 billion. Now, this round has concluded with over $3.5 billion raised, sending Moonshot AI's post-money valuation to $35 billion.

Looking back at this funding trajectory, the most dramatic changes have occurred in the past six months.

At the end of 2025, Moonshot AI's valuation was $4.3 billion. Just over half a year later, this figure has reached $35 billion, multiplying over seven times. If the next round is completed at a $50 billion pre-money valuation, Moonshot AI's valuation increase over just over three years will exceed 160-fold.

The reasoning from capital is also becoming increasingly direct. Previously, investors were betting on Yang Zhilin and a Tsinghua-affiliated technical team. Now, they are betting on whether Moonshot AI can become one of the few Chinese companies remaining at the global frontier model table.

After K3, Moonshot AI Returns to the Center of the Table

The early closure of this financing round perfectly coincides with the release of Kimi K3.

On July 16th, Moonshot AI released Kimi K3. On July 27th, the company further open-sourced K3's complete model weights, technical report, and part of its key infrastructure technology.

K3 is a Mixture of Experts model with 2.8 trillion total parameters and 104 billion activated parameters per inference. It supports a 1 million token context window and possesses capabilities in vision, coding, reasoning, and long-term tool use.

Moonshot AI stated in its technical report that K3 achieves state-of-the-art performance on tasks involving long-context programming, Agents, knowledge, reasoning, and vision. Its overall capabilities are very close to the strongest closed-source models, Claude Fable 5 and GPT-5.6 Sol.

Leaderboards change constantly, and one must be cautious of AI model companies' own benchmarks. K3's more significant impact is pushing the capability boundaries of open-weight models forward once again.

For a long time, open-source models primarily played catch-up. Closed-source companies trained the most powerful models, and six months or a year later, the open-source community would reproduce some of those capabilities at a lower cost.

K3 changes this time lag. A Chinese company has directly released a model with capabilities approaching the global frontier. Other developers can download, modify, deploy it, and even continue training their own models based on it.

Of course, this doesn't mean the model has become cheap. K3's weight files exceed 1.5 TB. Even loading the complete model requires multiple high-end accelerator cards. Deploying it for production environments will likely involve significant hardware investment. Its "open" nature primarily addresses whether enterprises can control the model and avoid being locked into a single API provider; it doesn't mean everyone can run it on their home computer.

Even so, open-weight models inevitably impact the business of closed-source companies.

When a sufficiently capable model can be downloaded, enterprises have the opportunity to deploy it on their own servers, keep data in-house, customize the model, and avoid paying token prices dictated by a single company forever. The moats that closed-source models built on capability leadership will be eroded much faster.

Therefore, after K3's release, the debate quickly moved beyond technical leaderboards into the realm of policy and commercial interests within the US AI industry.

Behind the Open-Source vs. Closed-Source Debate

On July 24th, companies and institutions including OpenAI, Google, Microsoft, NVIDIA, AMD, Meta, and Hugging Face publicly supported open-weight models, opposing the US government's rush to restrict downloadable and deployable AI models.

Anthropic and Amazon were absent from the list of signatories. Subsequently, questions arose whether Anthropic was using the guise of safety and national security to limit the development of open models and block competitors from its own closed-source business.

On the very day K3's weights were released, Anthropic CEO Dario Amodei published an article responding to the controversy.

He stated that Anthropic never advocated for a complete ban on open-weight models. He argued that open models without dangerous capabilities are a public good, lowering costs and creating value for enterprises, developers, and researchers.

However, Amodei's core position remained unchanged.

He continued to advocate for restricting the flow of advanced chips and chip manufacturing equipment to China, cracking down on so-called "industrial-scale model distillation," and requiring mandatory safety testing before releasing sufficiently capable models, regardless of whether they are open or closed source.

While the debate revolves around safety on the surface, there is a very tangible business issue underneath.

Closed-source companies want to maintain control over the most powerful models, charging fees via APIs and subscriptions. Open models, conversely, diffuse capabilities to more companies, lower prices, and shorten the distance between latecomers and leading firms. The faster models are opened, the harder it is for the capability advantages to be locked away long-term within a few companies' servers.

K3 comes from China, is massive in scale, and its capabilities are near the frontier. For proponents of open models, K3 proves that the open-source route can still produce world-class models. For Anthropic, it also means that the capability advantages built by US companies at enormous cost could proliferate much faster.

This has granted Moonshot AI a new kind of capital value it previously lacked.

It is no longer just an AI application company with a large number of Chinese users, nor just a foundational model startup waiting for its business model to mature. After the release of K3, it began to be re-evaluated within the context of multiple competitions: global open-source vs. closed-source, US vs. China, model capability vs. computing power control.

The $3.5 billion funding round and $35 billion valuation are the answers provided by the capital market.

Original Article Link