4 more excellent local LLM projects you can run for free on a slow laptop
msn.comGuide to running local LLMs on slow laptops with limited RAM, showcasing 4 free projects that enable AI inference without cloud dependency.
The Year of the Claw: OpenClaw Rolls Out New Version of Its Viral AI Agent
msn.comOpenClaw rolls out version 2.0 of its viral AI agent platform, marking a significant update to the open-source agentic system with new features and improvements.
Genians Finds Kimsuky Used OpenCode in GitHub PAT-Backed LNK Campaign
letsdatascience.comSouth Korean security firm Genians analyzed malicious LNK files and found opencode (an AI coding assistant) embedded in metadata of decoy PDFs used by Kimsuky threat actors.
LMArena 評比出爐：GPT-6 Astra 稱霸 Code Arena，Fable 5.1 在 Intelligence Index 仍守第一
inside.com.twLMArena benchmarks show GPT-6 Astra dominates Code Arena while Fable 5.1 maintains top position on Intelligence Index.
OpenAI touts new model Astra as 'world's most intelligent'
msn.comCNBC's Andrew Ross Sorkin reports on OpenAI touting its new model Astra as the world's most intelligent, highlighting claims about AGI arrival.
Sam Altman apologizes for 'messy' GPT-6 Astra rollout that's locked out paying users
theverge.comOpenAI CEO Sam Altman apologized for the messy rollout of GPT-6 Astra, which has locked out paying users despite being marketed as a major new model.
Cohere Health Recognized in 2026 Gartner® Hype Cycle™ for U.S. Healthcare Payers for Fifth Consecutive Year
uk.finance.yahoo.comCohere Health, a clinical intelligence platform for health plans, was recognized in Gartner's Hype Cycle for U.S. Healthcare Payers for the fifth consecutive year (note: this is about healthcare AI applications).
Mistral AI开放生态引爆2026全球AI主权浪潮！GLM-5.2首登欧洲平台
sohu.comMistral AI opened its platform to global third-party open-source models, with Zhipu AI's GLM-5.2 being the first to access this European AI gateway.
NVIDIA has unveiled the 'Nemotron 3.5 Lightning,' a high-speed model for AI agents, and also introduced 'NeMo Switchyard'
gigazine.netNVIDIA announced Nemotron 3.5 Lightning, an open model for AI agents, and NeMo Switchyard, a tool that automatically selects the most suitable AI model.
Everything we know about Z.ai, the Chinese company behind the mysterious Ox Alpha model
msn.comZ.ai revealed that mystery AI model Ox Alpha was GLM-5.3-Flash, which it had anonymously tested on OpenRouter and OpenCode. The article covers the Chinese company behind this record-breaking open-weight model released under a different name initially.
China's GLM-5.3-Flash Just Beat Claude Opus 4.8 on Agent Benchmarks
startupfortune.comZhipu AI's GLM-5.3-Flash beats Claude Opus 4.8 on agent benchmarks at 1/42nd the price, with MIT-licensed weights now live. The article discusses this Chinese model outperforming US systems in cost-effective performance for autonomous agents.
DeepSeek Plans a Massive Huawei AI Chip Deployment
techgenyz.comChinese AI lab DeepSeek is pushing for massive deployment of Huawei Ascend 950DT accelerators, targeting a new data center in Inner Mongolia. The move represents strategic shift toward domestic AI chip alternatives amid US export restrictions on NVIDIA hardware.
DeepSeek reportedly plans 160,000-chip Huawei cluster in Inner Mongolia
technode.comDeepSeek is reportedly planning to deploy at least 160,000 Huawei Ascend 950DT accelerators in a new AI data center being built in Inner Mongolia. The infrastructure investment represents significant commitment to domestic chip alternatives for training and inference workloads.
Minnesota's AI 'Nudification' Law Already Changed Grok — Now xAI Has Lost Its First Court Bid
eweek.comxAI's first court bid to challenge Minnesota's AI 'nudification' law was lost, leaving Grok restrictions in place as the company appeals and continues legal challenges against state regulations affecting its product.
AWS Opens GPT-5.6 Access on Amazon Bedrock From Australian Regions
unite.aiAmazon Web Services announced that teams in Australia can now access OpenAI's GPT-5.6 models through Amazon Bedrock from Australian regions starting September 2, 2026.
Claude Can Help Manage Your Email Inbox, But There Are Some Risks Involved
aol.comAnthropic's Claude can now fully manage Gmail inboxes, including sending and replying to emails. The article discusses the capabilities of this AI feature while noting potential risks involved with using an AI assistant for email management.
ChatGPT, Grok, and Claude all went down at the same time
theverge.comMajor AI chatbots including ChatGPT, Grok, and Claude experienced simultaneous outages around 11AM ET on Thursday morning. Users reported widespread access issues with these platforms before services were restored later that day.
Nvidia Groq 3 LPX enters full production for faster agentic AI inference
digitimes.comNvidia announced its Groq 3 LPX inference accelerator is now in full production, which could speed up agentic AI systems used worldwide for coding, reasoning, and other tasks.
China's Moonshot AI in negotiations with Microsoft, Amazon, Google over Kimi K3 revenue sharing
msn.comIPO-bound Moonshot AI is in negotiations with Microsoft, Amazon and Google over revenue sharing for their Kimi K3 model. The company wants up to 30% of the revenue generated from services offered through these hyperscalers' cloud platforms while establishing US enterprise distribution channels.
US hyperscalers discuss revenue share with China's Moonshot K3 model: Reuters
blockspace.mediaMoonshot AI is seeking up to 30% of revenue from Kimi K3-related services on Microsoft Azure, Amazon AWS, and Google Cloud. The proposed agreements would give Moonshot access to US enterprise distribution while requiring hyperscalers to operate a specific infrastructure setup for the Chinese model provider.