DishwasherBench introduces physical simulation for LLM spatial reasoning
October 13, 2026
DishwasherBench is a 3D rigid-body physics simulation designed to benchmark the spatial reasoning and planning capabilities of language models via Model Context Protocol (MCP). The environment requires models to optimize the loading of a Bosch dishwasher to ensure efficient cleaning.
HOW THIS AFFECTS YOU
●
researcherYou can use this deterministic simulation to evaluate how well models handle complex physical constraints and spatial planning.