Alibaba-backed Chinese artificial intelligence startup Moonshot just unveiled its latest model, Kimi K3, and it’s already sending shockwaves through the industry, with some benchmarks showing the model outperforming Anthropic and OpenAI’s best offerings.

The model packs 2.8 trillion parameters, which Moonshot says would make it the largest open-weight model released to date once its weights become available by July 27.

In a blog post, the company acknowledged that K3’s overall performance still trails Claude Fable 5 and GPT-5.6 Sol. Its internal evaluations nevertheless place it close to both models on several tasks, while independent testing by Artificial Analysis ranks it immediately behind the leading proprietary systems on its Intelligence Index and real-world work evaluations.

On Arena.ai’s front-end development leaderboard, K3 even ranks above the two most powerful models, marking a 17-place jump from the company’s previous model, Kimi K2.6. Arena’s CEO, Anastasios Angelopoulos, said Kimi K3 “may be the single biggest release of the year” and “the moment that OSS Chinese models have surpassed US models,” in a post on X.

Anastasios Nikolas Angelopoulos (@ml_angelopoulos) [https://xcancel.com/ml_angelopoulos/status/2077832882673066109]

This may be the single biggest release of the year, and marks the moment that OSS Chinese modles have surpassed US models.

Code Arena, Kimi K3 has BEATEN FABLE.

This is only 6 weeks after the Fable release.

This makes @Kimi_Moonshot the #1 AI lab in the world on frontend coding capability, and more results are rolling in that are likely to continue to show it is at the top of the pack.

The implications of this, whether on the closed-source AI business models or the larger capital ecosystem in the US, are enormous.

Arena.ai (@arena) [https://xcancel.com/arena/status/2077824029126504525]

Big news: Kimi-K3 by @Kimi_Moonshot is now #1 in the Frontend Code Arena with 1679 pts, surpassing Claude Fable 5.

This is a 17-place jump from Kimi-k2.6 (#18 -> #1).

In Frontend, Kimi-K3 ranked #1 in 6 of 7 domains: Brand & Marketing, Reference-Based Design, Data & Analytics, Consumer Product, Simulations, and Content Creation Tools, landing #2 only in Gaming behind Fable 5.

The full model weights will be released by July 27.

Congrats to the @Kimi_Moonshot team on this major milestone!

It’s a remarkable achievement, especially for an open-source model. The results challenge the assumption that China’s leading AI labs remain several months behind their American competitors. Anthropic just released Fable 5 last month, while OpenAI’s GPT-5.6 (and its three tiers, Sol, Terra, and Luna) just dropped last week.

“Kimi k3 is a big moment with multiple implications for the entire industry,” Trump’s former senior White House policy advisor on AI, Sriram Krishnan, said in a post on X.

The last time something like this happened, aka when a Chinese AI lab released a cheaper model that proved competitive with American alternatives, was when DeepSeek released R1 back in January 2025. Following that release and its reception, the market reaction helped wipe roughly $1 trillion from global technology stocks. Meanwhile, the model’s success raised major national security concerns across Washington D.C., and partially informed the Trump administration’s hard-line stance on advanced tech exports to China.

Moonshot’s release also comes only a few months after Anthropic accused the company, along with other Chinese AI companies DeepSeek and MiniMax, of violating their rules to “illicitly” extract the capabilities of its model Claude and use that to improve their own models. The process is called “distillation,” and it’s fairly common in the industry, but the Trump administration has deemed it “adversarial” and vowed to crack down on it.

K3 arrives amid heightened scrutiny of the U.S.-China AI race and growing national-security concerns around frontier models. Its release is likely to renew debate in Washington over export controls, distillation, and whether restrictions on Chinese labs are slowing their progress at all.

Spy Family

The Central Intelligence Agency (CIA) has been stepping up its efforts in the world of AI — including an eyebrow-raising use of chatbot tech.

As the New York Times reports, the CIA has been quietly developing a platform that lets analysts “talk” to foreign leaders, in a bid to predict how they might react in certain situations. The human variety of this type of behavior-predicting analysis has been the bread and butter of the agency’s behind-the-scenes grunts for a very long time. Instead of painstakingly compiling “profiles” on world leaders based on public information and gathered intelligence, however, those analysts will engage in faux conversation with large language models (LLMs) trained on similar intelligence and information that’s presumably being fed into its training data.

The NYT didn’t say how formally the chatbot has been deployed, or who helped develop it. However, an interview with the CIA’s first chief technology officer, ex-Pentagon AI czar Nand Mulchandani, reveals that its opacity is very much by design. Mulchadani, a Silicon Valley veteran, has a chart in his offices showing all the layers of approval it takes to get any private sector collaboration approved within the secretive agency. From handing issues with contracts to taking care of any project roadblocks, each step requires an incredible amount of bureaucracy and clandestine discussion — hurdles that the CIA acknowledges are hindering its quest to keep up with innovation and China, America’s main tech adversary.

Training Day

The agency’s now-CTO was, as the NYT notes, hired to help spearhead a forward-thinking sea change within the CIA. In the two-and-a-half years since he was brought on, Mulchadani has apparently made it easier for private companies to start working with the intelligence agency — and reading between the lines, it seems he’s held the hands of tech CEOs through the labyrinthine bureaucracy.

“The more we share about how we employ technology, how we procure technology, what we’re going to do with it, will make companies want to work with us and want to team with us more,” explained Juliane Gallina, the deputy director of the CIA’s digital innovation arm, in an interview with the NYT.

According to Gallina, the agency is looking to declassify and “expose a little bit” of its secret technological sauce to help procure private sector contracts.

There was no mention, however, of whether the public will be given a look behind the curtain of what their tax dollars are helping to fund.

More on spies: Hackers Apparently Stole the FBI’s Call Logs With Confidential Informants