The macOS beta is on its way — every download below goes live with it. Drop your email and we'll send the release announcement. Product updates only, unsubscribe anytime.
The full control platform: jog, 3D toolpath, part nesting, mesh mapping, job cycles, macros, vision safety, MCP for AI agents — and a built-in simulator, so you can run all of it before wiring a single machine.
The companion that lives next to your controller: it owns the USB serial link and the robust streaming loop, plus the second serial port for the IO board. Your Mac — and your AI — drive it over LAN or VPN.
Firmware for the IO board: relay outputs for vacuum, lights, clamps, and mist, plus debounced fault inputs — driver alarms, door switches, e-stop — that can halt the machine on their own.
One-command setup scripts that install llama.cpp and pull the vision model that powers setup checks and camera questions. Runs entirely on your Mac — any OpenAI-compatible vision endpoint works too.
Working examples of automated job cycles: load-and-clamp scripts with vision checks, arm flip and re-clamp sequences, IO routines, and sample programs to run against the simulator on day one.
The safety-first path from unboxing to first cut: serial link, first motion, remote mode, IO wiring, alarm inputs, camera — each phase with explicit pass criteria before you move to the next.