VideoZeroBench: Probing the Limits of Video MLLMs with Spatio-Temporal Evidence Verification

TL;DR

VideoZeroBench is a new hierarchical benchmark for long-video question answering that rigorously verifies spatio-temporal evidence, exposing gaps in current models' grounded video understanding capabilities.

Contribution

It introduces a challenging, multi-level evaluation protocol and a comprehensive dataset with annotated spatio-temporal evidence for long-video QA.

Findings

Gemini-3-Pro answers fewer than 17% of questions correctly under standard conditions.

Performance drops below 1% accuracy when strict evidence grounding is required.

Most models fail to produce correct grounded predictions, highlighting a gap in grounded video reasoning.

Abstract

Recent video multimodal large language models achieve impressive results across various benchmarks. However, current evaluations suffer from two critical limitations: (1) inflated scores can mask deficiencies in fine-grained visual understanding and reasoning, and (2) answer correctness is often measured without verifying whether models identify the precise spatio-temporal evidence supporting their predictions. To address this, we present VideoZeroBench, a hierarchical benchmark designed for challenging long-video question answering that rigorously verifies spatio-temporal evidence. It comprises 500 manually annotated questions across 13 domains, paired with temporal intervals and spatial bounding boxes as evidence. To disentangle answering generation, temporal grounding, and spatial grounding, we introduce a five-level evaluation protocol that progressively tightens evidence requirements. Experiments show that even Gemini-3-Pro correctly answers fewer than 17% of questions under the standard end-to-end QA setting (Level-3). When grounding constraints are imposed, performance drops sharply: No model exceeds 1% accuracy when both correct answering and accurate spatio-temporal localization are required (Level-5), with most failing to achieve any correct grounded predictions. These results expose a significant gap between surface-level answer correctness and genuine evidence-based reasoning, revealing that grounded video understanding remains a bottleneck for long-video QA. We further analyze performance across minimal evidence spans, atomic abilities, and inference paradigms, providing insights for future research in grounded video reasoning. The benchmark and code will be made publicly available.