Benchmark RadarAI BENCHMARK PROFILE
MECoBench
MECoBench is a multimodal embodied cooperation benchmark with an evaluation platform. It spans real-world tasks, two cooperation structures, and three collaboration modes, with code and dataset publicly available.
- Released
- 2026-06-30
- Readiness
- Runnable
- Primary field
- Robotics & Autonomous Systems
Why it matters
Systematically evaluates collaboration among multimodal embodied agents, a relatively unexplored area. Provides a testbed for understanding collaboration mechanisms and limits.
Motivation
Recent multimodal large language models (MLLMs) have strong potential as embodied agents, but their ability to collaborate in visually grounded environments remains underexplored.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.