Apple Inc. announced new Mac Mini and Mac Studio desktops with M6, M5 Pro, M5 Max, and M5 Ultra chips, featuring a 2nm process and a doubled Neural Engine (32 cores) for AI performance, with prices ri…
Apple Inc. launched new Mac Mini and Mac Studio desktops on Tuesday, featuring the M6 and M5 Pro chips in the Mac Mini and M5 Max and M5 Ultra in the Mac Studio, with performance gains aimed at AI dev…
Apple unveiled the next-generation Mac Studio, powered by the M5 Max and M5 Ultra chips, with the M5 Ultra offering up to a 36-core CPU, an 80-core GPU, and up to 512GB of unified memory. The M5 Max M…
A developer has released a CUDA-optimized fork of antirez's h3.c for NVIDIA DGX Spark, achieving a ~15.5x speedup, while also providing a native Apple Silicon implementation for MiniMax-H3 inference w…
Meta released Muse Glimmer, a 30-billion-parameter multimodal agent designed to run on consumer hardware with 24 GB or 32 GB of memory, under the Apache 2.0 license on August 10. Tomas Koutsky, co-fou…
Meta released Muse Glimmer 30B on August 10, 2026, an open-weight model under Apache 2.0 designed for local AI agents, fitting on a single consumer GPU. It reached #1 on Hacker News with over 1,000 po…
Antirez released h3.c, a native MiniMax-H3 inference engine for Apple Silicon Macs, with prompt-to-video/audio, first/last-frame conditioning, and ordered Ref2VA references working end to end. The pro…
Meta released Muse Glimmer, a 30-billion-parameter open-weight model for local AI agents, available now on HuggingFace under Apache 2.0. The model, from Meta Superintelligence Labs, targets developers…
In 2026, edge inference has advanced to the point where an 80B Qwen model runs in just 4.3GB of RAM on a MacBook Pro, and a 35B model runs on an iPhone 18 Pro. This is achieved through a combination o…
A 60-run evaluation of coding agents found no quality winner among local DeepSeek V4 Flash, hosted DeepSeek via Vercel AI Gateway, and Claude Sonnet 5, with strict acceptance at 8/20, 9/20, and 5/20 r…
Apple's next Mac Studio, expected to feature M5 Max and M5 Ultra chips, may be delayed until October 2026 due to global DRAM shortages driven by surging AI infrastructure demand, according to Bloomber…
Ollama 0.31 introduces multi-token prediction for Gemma 4 on Apple Silicon, cutting token generation time by 90% on Aider's coding benchmark. The update automatically accelerates local coding agents w…
Small language models (SLMs) are increasingly powering next-generation AI agents, challenging the assumption that larger models are always better. NVIDIA research shows SLMs excel at repetitive, speci…
Apple's upcoming 'MacBook Ultra' will use M5 Pro and M5 Max chips, disappointing some users. The iPhone 18 and iPhone 18e are rumored to feature 9GB RAM and an A20 chip. A cyberattack at Apple supplie…
Apple is developing a high-end MacBook Ultra with an OLED touchscreen display, Dynamic Island notch replacement, and M5 Pro/M5 Max chips. The device, expected in 14-inch and 16-inch sizes, marks the f…
Apple is reportedly working on an M7 Ultra Mac Studio for 2028, according to Bloomberg's Mark Gurman, while an M5-based Mac Studio refresh is still on track for later this year. The current Mac Studio…
Apple's upcoming touchscreen MacBook Pro will launch with M5 Pro and M5 Max chips instead of the previously expected M6 series, according to Bloomberg's Mark Gurman. The 14-inch and 16-inch laptops wi…
A developer building a coding agent called ForgeFlow discovered that their pytest test suite reported 186 passed tests even though the tests were running in the wrong Python virtual environment. The e…
Apple plans to launch four new Mac models later this year, including a Mac Studio with M5 Max and M5 Ultra chips, an M5 iMac with a refreshed color palette, and a Mac mini with M5 and M5 Pro options. …
Local AI users are accumulating hundreds of gigabytes of model files, including multiple quantizations of the same model, image and video generation weights, and RAG datasets, often with duplicates th…