openreality

Open Reality turns a phone video into a persistent 3D scene that an AI agent can query through 41 MCP tools. It supports measurement, path planning and robot-training exports, and the whole server can be self-hosted on your own GPU or Modal account.

openreality sits in the Vision / Multimodal category, so it is usually better judged by workflow fit, deployment style, and team needs than by isolated features alone. Based on the available metadata, it leans most clearly toward Upload a phone video and get a persistent 3D scene a few minutes later. A representative use case is Ask Claude Code how big a room is or what objects are in a scan. The listed license is BSD-2-Clause. GitHub stars are currently 424. Relevant tags include 3d, 3d-reconstruction.