Benchmark RadarVideoGAIA
VideoGAIA evaluates agentic video understanding through multi-turn, tool-augmented interactions where models must iteratively perceive videos, invoke external tools, and integrate multimodal evidence. It contains 271 human-verified tasks across diverse real-world scenarios.
- Released
- 2026-08-12
- Readiness
- Runnable
- Primary field
- General AI
Why it matters
Conventional single-turn video understanding benchmarks are becoming saturated; VideoGAIA moves beyond to assess advanced MLLMs' ability to act as general AI assistants, using tools to gather complementary information across turns. It provides a timely, challenging benchmark for next-generation video understanding.
Motivation
Video understanding is a fundamental task for evaluating the capabilities of multimodal large language models (MLLMs).
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.