Studio Sentinel
Watch your Mac think, from your phone
Start a long local-inference job on a Mac Mini or Mac Studio, walk away, and you normally have no idea what happened. Studio Sentinel puts the answer on your phone — live, over your own network, with nothing sent to a cloud.
CPU, core by core
Real-time usage across performance and efficiency cores, so you can see what inference is taking versus everything else.
Unified memory and pressure
Usage, pressure level (nominal, warning, critical), wired pages and free memory — the numbers that decide whether another model fits.
GPU and thermals
Apple Silicon GPU utilisation and unified-memory allocation, plus thermal state, so throttling stops being a mystery.
Built for local LLM servers
Ollama, LM Studio, llama.cpp and anything else you run on Apple Silicon.
Stays on your network
No cloud relay and no subscription for the core features. The data never leaves your LAN.
Screens
