Local Inference
The latest Local Inference coverage — news, analysis, and updates from the WindowsNews.AI desk.
Copilot App Gets On-Device Semantic Search: Find Files by Meaning, Not Name
Microsoft began rolling out an update to the Copilot app for Windows Insiders that introduces on-device semantic file search, allowing users to find documents and images by describing them in natural...
Microsoft's Copilot App Now Understands 'Chicken Tostada Recipe' — Semantic Search Hits Windows Insiders
Microsoft began rolling out a significant update to the Copilot app in Windows 11 on August 20, 2025, introducing semantic file search and a redesigned home surface to Windows Insiders with Copilot+...
Leading GenAI Browser Assistants Found Hoovering Up Social Security Numbers, Medical Data Without Consent
A sweeping new audit of the most popular generative AI browser assistants reveals that several widely used extensions silently capture and transmit highly sensitive information from ordinary web...
OpenAI's Open-Weight Model Flunks a 10-Year-Old's Exam Despite Solid Reasoning
OpenAI's latest open-weight model, gpt-oss-20b, demonstrated strong chain-of-thought reasoning when faced with a real UK 11+ exam—but its final answers were so wrong that an actual 10-year-old...
Windows 11 Now Runs OpenAI's GPT-OSS-20B Locally; Tech Jobs and Government IT Brace for Impact
Microsoft has quietly kicked open the door to a new era of PC-based AI, integrating OpenAI’s first open-weight model since GPT‑2 directly into Windows 11. The move lands in the same week that...
Microsoft Plants Open-Weight GPT Flag on Windows and Azure, Handing Developers Full Model Keys
Microsoft dropped a bombshell on the AI landscape this week, delivering what developers have craved for years: unrestricted access to the inner workings of advanced language models. The company is...
GPU-Accelerated gpt-oss-20B: Revolutionizing Generative AI on Windows with Local Inference
A transformative leap has arrived in the evolution of Windows-based artificial intelligence: the introduction of the GPU-accelerated gpt-oss-20B model. This development marks a substantial shift,...
OpenAI Releases Open-Weight GPT-OSS 120B & 20B Models Under Apache 2.0 License
A transformative moment is unfolding in the world of artificial intelligence as OpenAI officially launches its open-weight language models, gpt-oss-120b and gpt-oss-20b. These releases don’t just...
On-Device Mu AI in Windows 11 Cuts Latency to Under 100ms
Microsoft is rewriting the rules of on-device artificial intelligence with its groundbreaking Mu language model, a compact yet powerful AI solution designed specifically for Windows 11. This...
Setting Up a Local AI Chatbot on Windows Using PowerToys and Ollama
Introduction Artificial intelligence (AI) is increasingly becoming a staple in our daily computing experiences. However, concerns about data privacy and reliance on cloud services have led many users...
Windows 11 Gets DeepSeek R1 AI for Faster On-Device Processing
Microsoft has unveiled DeepSeek R1, a cutting-edge AI model designed to revolutionize the Windows 11 experience through local inference and neural processing. This launch marks a significant step in...