The Kimi K3 Storm Shakes Silicon Valley: OpenAI and Anthropic Lose More Than $300 Billion in Value

The cards are being reshuffled in the AI ecosystem, and this time the shock hit Silicon Valley's biggest giants directly. Kimi K3, the open-weight AI model released by China-based Moonshot AI, landed like an earthquake in the AI world thanks to its performance and high cost efficiency.

Kimi K3's ability to go head to head with the most advanced closed-source Western models was not just a technological achievement; it also led investors to question the enormous valuations they had placed on giants such as OpenAI and Anthropic.

A $314 Billion Valuation Earthquake in the Markets

Kimi K3's arrival shook the perception of "the overwhelming superiority of closed models" in the AI industry to its core. Well-known market analyst Tony Sycamore, speaking to Bloomberg, said that within just a few days of Kimi K3's release, more than $314 billion had been wiped off the combined valuation of OpenAI and Anthropic.

So how did the release of an open model cause such a massive financial collapse?

1. Loss of Pricing Power: Companies like OpenAI and Anthropic offered their closed models at high API and subscription prices to cover their high development and infrastructure costs. High-performance open-weight models like Kimi K3 let companies and developers access a similar level of AI capability at a much lower cost. 2. Weakening Economic Moats: Market experts and investors believed there was an insurmountable "technological moat" around the giants that own closed-source models. Kimi K3 crossing that moat put the long-term profitability expectations and monopoly positions of closed-model developers at risk. 3. Shifting Enterprise Preferences: As privacy- and data-security-focused enterprise customers began to prefer open models they can run on their own infrastructure or private clouds, the markets priced in the fear that closed-model providers would lose customers.

A Giant 2.8 Trillion Parameter Architecture: Kimi K3's Key Features

The impact Kimi K3 made is no accident. The model represents one of the most advanced levels of engineering reached in AI architecture:

* 2.8 Trillion Parameters: Parameter count is one of the most critical indicators of the depth of knowledge a model is trained on and its ability to form logical relationships. With its 2.8 trillion parameter scale, Kimi K3 sits at the top among the largest and most capable models in the open-source world. * 1 Million Token Context Window: Within a single prompt, the model can read and keep in memory a huge dataset of roughly 750,000 to 800,000 words. This makes it possible to analyze large codebases, complete legal files or thick technical documentation in one go. * Long-Running Autonomous Tasks: Kimi K3 has been specifically optimized for complex reasoning, multi-step software development projects and knowledge-intensive professional work.

Heavy Demand Locked Up the Infrastructure: Moonshot AI Paused Subscriptions

The capabilities Kimi K3 offers and the huge excitement it generated led to an unprecedented rush of users at Moonshot AI. However, this intense demand created a serious bottleneck in the company's cloud inference infrastructure.

With server capacity falling short of the excessive user demand, Moonshot AI had to temporarily stop accepting new paid subscriptions to protect the experience quality of existing users and prevent system latency.

On one side, hundreds of billions of dollars in lost value at OpenAI and Anthropic; on the other, Moonshot AI unable to keep up with demand and closing its doors. Together they clearly show the scale of the shock Kimi K3 caused in the AI market.

A New Era in the AI Economy

The Kimi K3 move was one of the harshest responses from the East to the West's claim to leadership in the AI race. Open-source and open-weight models catching up with, and in some areas even surpassing, closed systems protected by billions of dollars of investment are forcing the industry to fundamentally change its business models.

Everyone is now watching to see how giants like OpenAI and Anthropic will respond to this competition, whether they will change their pricing strategies, and how they will prove again the added value that closed models create.