Liquid AI and Artificial Analysis launch Pipette, an open-source benchmarking platform testing 30+ AI models across 1,000+ edge device configurations, delivering real-world performance data on devices like the iPhone 17 Pro and MacBook Pro M5 Max through native apps and an interactive public dashboard.
Chinese state-affiliated hackers have more than doubled their cyberattack volume by integrating DeepSeek and other open-source AI models into their operations, using the technology to automate routine tasks and develop advanced malicious software, according to Taiwanese research firm TeamT5.
Hugging Face is exploring a sale at a staggering $13 billion valuation — nearly triple its 2023 worth — as the AI developer platform, home to over 3 million public models, works with a bank to gauge buyer interest amid surging demand for AI middleware acquisitions.
A powerful open-source AI quantitative trading platform launches on GitHub, bringing LLM-driven strategy generation, real-time stock monitoring, and backtesting tools to A-share market retail investors and quant enthusiasts via a modern FastAPI and React 18 stack.
A free, open-source platform called VoiceStudio is challenging ElevenLabs by offering voice cloning, video dubbing, and transcription across 646 languages with no account or subscription required, running entirely on-device across Mac, Windows, and Linux.
Google's open AI model family Gemma surpasses one billion downloads as developers publish over 100,000 model variants, powering real-world breakthroughs in space exploration with NASA, cancer research with Yale, health data management for 100 million users in India, and dolphin communication studies with Georgia Tech.
Block open sources Berd, a desktop app that unifies fragmented AI agent workflows across tools and models, featuring animated 'Gloopies' characters to make agent configurations more approachable, while its design principles now influence Buzz, Block's new collaborative human-agent workspace.
Open-source AI tool dot-skill has surpassed 23,000 GitHub stars as it enables users to transform real colleagues, family members, celebrities, and even fictional characters into interactive AI Skills using chat logs, documents, and interviews across multiple agent platforms.
A new GitHub workshop, 'Agents That Own Their Inference,' teaches engineers to build production AI agents on dedicated GPUs using vLLM, covering nine modules on inference optimization and Kubernetes deployment with full open-source control and no vendor lock-in, available for roughly $0.52 per hour on Akamai Cloud.
Alibaba's new Qwen 3.8 27B vision AI model runs on consumer laptops as a 17GB file, delivering impressive code generation and image analysis capabilities, but its default 'xhigh' reasoning setting causes excessive overthinking on simple tasks, and speed remains limited at 15–30 tokens per second — though a 72% boost …