All articles tagged with #kimi k3
Moonshot AI debuts Kimi K3, a 2.8-trillion-parameter model it says narrows the gap with OpenAI and Anthropic, outperforming other tested models on benchmarks like coding and agents but still behind Claude Fable 5 and GPT 5.6 Sol on overall performance; the release signals China's tightening push in the AI race, despite compute constraints, and has drawn attention from investors and analysts alike.
Global markets fell after Moonshot unveiled Kimi K3, a powerful Chinese AI model that narrows the gap with OpenAI and Anthropic, fueling fears that open‑source AI could disrupt US incumbents and slow the AI spending rally. The Nasdaq dropped about 1.5% on Friday (S&P 500 -0.7%; Dow -60 points). Asia’s markets also fell (Taiwan down >6%, Japan about -4%). Micron slid roughly 30% from late June highs, while Alphabet and Nvidia edged lower. Oil rose as concerns over inflation persist; US gas neared $4 per gallon and June CPI showed inflation at 3.5% year over year. Investors are weighing how impactful Kimi K3 will be and whether AI hype will sustain chipmakers’ growth.
Semiconductor stocks fell as investors adopted a risk-off stance, with MU, SNDK, and WDC among the losers and Nvidia, AMD, Broadcom, Intel, and Marvell also slipping; Applied Materials and Lam Research declined more than 4% as the sector’s value shrank by about $3.3 trillion since June 22. The move comes as AI infrastructure costs rise and hyperscalers’ capex remains a key driver, while Moonshot unveiled Kimi K3—the world’s largest publicly available AI model—at WAIC in Shanghai, highlighting China’s rising open AI push. Alphabet’s Gemini 3.5 Pro reportedly remains behind schedule, underscoring mixed momentum in the AI race even as spending persists.
Moonshot's Kimi K3, a 2.8 trillion-parameter open-weight model with a 1-million-token context, reportedly beats leading U.S. models in coding benchmarks and outperforms or ties frontier rivals in Arena tests; priced around $12 per million tokens and with weights due July 27, the release signals cheaper, broader access to frontier AI and intensifies the U.S.–China AI competition, though early results may overstate real-world performance.