Open source evals for physical AI. Run any LLM/VLA on any arm/humanoid against any real/sim benchmark.
SDK for robotics teams to verify the quality of their data used for AI model training.
Open source evals for physical AI. Run any LLM/VLA on any arm/humanoid against any real/sim benchmark.
SDK for robotics teams to verify the quality of their data used for AI model training.