Open source evals for physical AI. Run any LLM/VLA on any arm/humanoid against any real/sim benchmark.
Open on GitHub