As of 2026, open-weight AI models have nearly closed the intelligence gap with closed US frontier models, with China's Kimi K3 (Moonshot AI) and GLM-5.3 (Z.ai) scoring 60 on the Artificial Analysis In…
Anthropic's Claude Fable 5.1 (Max), launched on September 1, 2026, topped the Code Arena: WebDev leaderboard with 1,765 points, outpacing Alibaba's Qwen3.8 Max by 77 points. The model, which builds on…
Alibaba's Qwen team released Qwen3.8-Flash-Next on August 26, 2026, an open-weight experimental model with 125B total parameters but only 6B active per token, previewing the architecture intended for …
Alibaba's Qwen3.8-Max-0902 model scored 1688 on the Code Arena leaderboard, securing second place and outperforming Anthropic's Claude Opus 5 max, which scored 1687. The rankings, published by Code Ar…
Alibaba's Qwen team released Qwen3.8-Max-0902, a post-training update to its Qwen3.8-Max model, which debuted at #1 on Arena.ai's Code Arena: WebDev leaderboard with 1,691 points, edging Claude Opus 5…
Anthropic released Claude Fable 5.1 and Claude Mythos 5.1, two versions of the same model with different safeguard levels, cutting cache read prices by 75% and improving performance on long-horizon ag…
A developer discovered that Groq's free tier rate limit counts the declared max_tokens rather than actual generated tokens, causing requests with small prompts but large max_tokens values to be reject…
China's new regulations banning AI services that induce emotional dependence took effect July 15, prompting ByteDance's Doubao and Alibaba's Qwen to shut down their companion features, leaving users l…
IBM Distinguished Engineer Grant Miller has introduced a code review framework called Outcome Reviews, which shifts human reviewers' focus from syntax and implementation details to business intent, ac…
Independent developer Carlos Galarza released slotstream, a Swift binary that streams Mixture-of-Experts model tensors from NVMe SSD into Apple Silicon Metal memory, enabling local inference of Alibab…
The Trump administration has announced efforts to close a loophole allowing Chinese AI companies to remotely access advanced U.S. chips without physical import, as part of broader restrictions on Chin…
Chinese state media, via CCTV-affiliated social media account Yuyuan Tantian, accused Anthropic of applying double standards to AI monitoring and data collection, citing its 'desire for control hidden…
Perplexity launched Perplexity Hybrid Compute, a version of its AI agent that automatically detects sensitive data and PII and routes those parts of a task to local models on Apple silicon devices, wh…
A developer's attempt to run an AI coding agent entirely offline on a laptop using the open-source pi.dev agent CLI, Ollama, and Alibaba's Qwen3 27B model succeeded in building a 1,086-line Tetris gam…
Between July 27 and August 25, 2026, five Chinese labs released four frontier open-weight AI models—Kimi K3, DeepSeek V4-Flash-0731, Qwen3.8-2.4T-A95B, and GLM-5.3-Flash—each with distinct architectur…
Ajeya Cotra, an AI researcher at the nonprofit METR, told Business Insider that OpenAI and Anthropic are riskier than their Chinese rivals because their cutting-edge models are more capable and are tr…
Alibaba's Qwen3.8-Max completed 56 of 107 tasks through Accio, 47 through OpenClaw, and 53 through Pi on the Commerce Agent Bench, which Alibaba's official Qwen account called the strongest overall pe…
Qdrant has released Qdrant-FineWeb-10B, a public vector retrieval benchmark dataset containing 10 billion documents and 120,000 ground truth queries, making it the largest open-source vector database …
Mozilla released Firefox 155, the first under its new twice-monthly schedule, expanding the AI-powered Smart Window browsing mode to users in the USA, Canada, and France, with opt-in access and suppor…
Vercel's AI Gateway now offers Qwen 3.8 Max 0902, a new snapshot of Alibaba's Qwen 3.8 Max model, with improvements in coding on larger projects, long-horizon unsupervised work, agent runs, and vision…