m-series

A fold animation for your MacBook, inspired by the iPhone Duo animation. With true customizable perspective, fold and snap style animations, hinge and snap sounds and vibrations, lock screen support, all stable at 120Hz.

ESPHome component for DALY H/K/M/S-Series and 100Balance BMS

Local web UI for h3.c — native Metal video/audio generation with MiniMax-H3 on Apple Silicon, no PyTorch or MLX required.

Tiny macOS CLI for short max-speed fan boosts on supported M-series Macs.

Galactus executes 744B and 235B MoE models on undersized Macs with bit‑perfect llama.cpp parity, RAM-as-cache execution, and a full local app featuring an agent, permission gate, code editor, authenticated server mode, scheduled unattended runs, and fully published measurements.

An easy to use notes app with a canvas

Interactive Apple Silicon runtime for NVIDIA SANA-WM — walk the camera with WASD, or run an LLM-driven adventure game with SANA visuals. PyTorch MPS, no CUDA. Companion to junafinity's full-pipeline renderer.

An agentic, self-correcting machine learning experimental loop built locally on Apple Silicon using MLX.

A lightweight macOS menu bar app combining clipboard history, system monitoring, sleep prevention, and mouse wheel reversal — one icon, one process, optimized for Apple Silicon.

Run DeepSeek-V4.1-Flash (551B MoE, IQ2_XXS) locally on Apple Silicon with 128K context, even though it's bigger than RAM. Patched llama.cpp Metal runtime streams experts from SSD with a GPU expert cache. Tested on M5 Pro 64 GB: 252 tok/s prefill, 16.3 tok/s decode.

Prism Sound ADA-8XR FireWire on Apple M-series Macs (Apple Silicon, macOS 26 Tahoe) via ASFireWire — verified recipe, tools & docs

A Yocto/OpenEmbedded BSP for Apple Silicon Macs

Troubleshooting USB Wi-Fi adapters in Kali Linux ARM64 virtualized on Apple Silicon M-series Macs.

Dependency-free macOS TUI temperature monitor for Apple Silicon with SoC, SSD/NAND, and battery readings, 60-minute trends, local history, and thermal alerts.

Docker-first, local-first AI workload toolkit for macOS Apple Silicon using Ollama, llama.cpp, LiteLLM, and Claude-compatible local endpoints.

Bare CUDA Metal: CUDA binaries and cuBLAS calls run bare-metal on Apple M-series GPUs — PTX→MSL translator, offline metal -c/metallib AOT into cached .metallib binaries, zero-copy unified memory, SIMD-aligned dispatch, bitwise-verified against NVIDIA CUDA. All live-verified.