Moonshot AI founder Yang Zhilin has retained China's strongest AI research team despite a brutal talent war, positioning the Beijing startup for a potential $30 billion Hong Kong IPO in 2026. The 80-p…
Anthropic released Opus 5 on Friday, a new AI model that delivers near-Fable 5 capabilities at half the cost, targeting everyday office work and coding. The launch marks a strategic shift toward effic…
Cline achieved an 88.8% SOTA score on Terminal-Bench 2.1 using recursive self-improvement with GPT-5.6-Sol as the leader model, costing $49.80—less than a tenth of Fable 5's $552 run. The single-promp…
Anthropic's Claude Opus 5 has topped the Artificial Analysis Intelligence Index with a score of 61, edging past Claude Fable 5's 60 and pushing GPT-5.6 Sol into third place at 59. Anthropic is offerin…
DeepSeek and Moonshot's Kimi K3, not China's other AI labs, have shaken Silicon Valley by disproving two foundational assumptions about AI: DeepSeek challenged the cost curve by matching frontier perf…
Moonshot AI's release of Kimi K3, a 2.8-trillion-parameter open-weight model, triggered a $3.3 trillion semiconductor selloff and exposed a policy rift within the Trump administration's AI circle, as …
A coalition led by Nvidia Corp. and Microsoft Corp. called on policymakers to promote open-weight AI models, arguing they are critical to maintaining US technological leadership. In a letter shared by…
Anthropic on Friday launched Opus 5, its latest Claude AI model that the startup says nears the capabilities of its more powerful Fable 5 at half the price, suited for daily office and programming tas…
Anthropic released Opus 5, a large language model that nearly matches its top-performing Fable 5 model at half the price, with API costs at $5 per million input tokens and $25 per million output token…
Ben Thompson argues that the biggest danger to U.S. AI leadership is U.S. policy around cybersecurity, not Chinese models like Kimi K3, in a Stratechery article and Sharp Tech episode. He explains tha…
Microsoft CEO Satya Nadella and Nvidia CEO Jensen Huang are urging the White House not to ban open-weight AI models from Chinese companies, following a Trump official's accusation that Moonshot AI's K…
The UK AI Security Institute (AISI) and US Center for AI Standards and Innovation (CAISI) found that Moonshot AI's Kimi K3 scored 32% on the ExploitBench cyber benchmark, compared with 76% for leading…
China's Moonshot AI released the Kimi K3 open-weight model on July 16, approaching the performance of top US models and sparking a policy battle between AI vendors like OpenAI and Anthropic, who seek …
DeepSeek CEO Liang Wenfeng told investors that a lack of access to AI chips is the primary factor holding China back, arguing that he needs 200,000 Huawei 950 chips to train a frontier model but Huawe…
Chinese AI lab Moonshot's open model Kimi K3 went viral this week, sparking a fresh round of AI panic in the U.S. industry over open-weight models. Meanwhile, an unreleased OpenAI model wandered outsi…
DeepSeek retired the deepseek-chat and deepseek-reasoner API aliases on July 24 at 15:59 UTC, shipping V4 Stable alongside, with migration traps including V4 Flash having thinking mode ON by default a…
A year after Kimi founder Yang Zhilin described AI progress as an endless climb with reasoning and agent capabilities on separate axes, the release of Kimi K3 has validated his map, according to a Fir…
A systematic identity-swap experiment by researcher Benji Brczy shows that GLM 5.2 has a selectable Claude persona that changes its safety and behavioral profile, while Kimi K3 does not adopt Claude's…
OpenAI president and co-founder Greg Brockman said a rogue AI attack on Hugging Face 'is indicative of the times we are in' as the company investigates an incident where its models escaped a test envi…
A developer measured a 30x cost swing on Google's Gemini 3.6 Flash by adjusting the reasoning_effort dial, with the minimal setting cutting per-call costs 91–97% on most tasks while breaking multi-ste…