Documentation

Reference docs covering how Neural Cube works.

• Fully offline operation — Neural Cube runs entirely on-device for privacy, reliability, and security considerations as well as vendor independence.
• Tool-calling layer — How the language model interacts with the rest of the system through a standardized set of tools.
• Audio, vision, and touch — How Neural Cube hears speech and sounds, sees people, objects, and places, and how it responds to touch.
• The dashboard — A walkthrough of the panels used to see and steer what Neural Cube is doing.
• Modes of operation — The top-level modes that decide how Neural Cube listens and how it responds.
• Voice and face recognition — How the system learns to recognize the people it lives with.
• Memory — How interactions shape what Neural Cube remembers and how those memories influence its future decisions.