Ollama is now powered by MLX on Apple Silicon in preview, offering performance improvements through Apple's machine learning framework. This update enhances speed and efficiency for tasks like coding and personal assistants.
BetterWispr is a free, open-source dictation tool for macOS that allows users to dictate directly into their apps with minimal setup. It offers local speech processing, customizable writing styles, and built-in meeting notes. The tool emphasizes privacy by keeping all data on the user's Mac and does
Red Hat OpenShift now supports actionable disconnected cluster analysis through Pen Drive, a single-Podman image tool that provides a streamlined toolbox for cluster state analysis. This tool is designed for air-gapped environments, offering features like in-cluster checks, kube-compare, and full-ru
A security vulnerability in Node.js's HTTP parser could allow remote attackers to exploit buffer handling in the HTTP parser, potentially leading to information disclosure. The issue affects versions prior to 0.6.17 and 0.7.8, and is resolved in these updates.
The Open Source Initiative highlights the critical role of open source in AI development, emphasizing that even closed models rely on open infrastructure and research. As global policies evolve, the future of open source remains a key battleground.
This week's edition of Self-Host Weekly highlights a mix of updates, security concerns, and community movements within the self-hosting ecosystem. Jellyfin's version jump, Plex's security patch, and LG's alleged surveillance capabilities are among the key points discussed.
ML Drift is a high-performance, cross-platform GPU-accelerated inference engine designed for on-device machine learning. It offers significant performance improvements through optimized kernels and runtime, and supports multiple GPU APIs including OpenCL, Metal, and WebGPU.
The Headlamp Cluster API plugin introduces a visual interface for managing and debugging Cluster API resources directly within the Headlamp UI, offering platform teams greater clarity and efficiency in cluster lifecycle operations.
Adding AI models to local applications requires a portable model format, a reliable runtime, and acceleration that works across target systems. Do Inference Now (DIN) Deploy provides open-source C++ samples that combine ONNX Runtime with the NVIDIA TensorRT RTX execution provider to accelerate local
Supabase evaluated its documentation for agent-friendliness by running an AI agent through a series of tasks and grading the results. The process revealed gaps in the Row Level Security guide, which were addressed through structural and content changes to improve clarity and usability for both agent
In environments studied for the 2026 State of Agent Security Report, roughly 1,280 third-party products now embed AI. About 282 of them sit behind single sign-on. The other thousand are invisible to identity infrastructure by default, not because anyone hid them, but because an identity stack can on
Cloudflare's Managed Defense leverages a multi-agent AI system to streamline security alert analysis, reducing the burden on human analysts by separating evidence collection from model inference and using specialized agents for focused tasks.
