SpatialBench: Benchmarking Multimodal Large Language Models for Spatial Cognition

TL;DR

This paper introduces SpatialBench, a comprehensive benchmark and hierarchical framework for evaluating the spatial cognition abilities of multimodal large language models across five levels of complexity.

Contribution

It proposes a hierarchical spatial cognition framework, constructs a detailed benchmark with 15 tasks, and introduces a unified metric for assessing spatial reasoning in MLLMs.

Findings

Models excel in perceptual grounding but struggle with symbolic reasoning.

Performance varies significantly across different cognitive levels.

Humans outperform models in goal-directed spatial abstraction.

Abstract

Spatial cognition is fundamental to real-world multimodal intelligence, allowing models to effectively interact with the physical environment. While multimodal large language models (MLLMs) have made significant strides, existing benchmarks often oversimplify spatial cognition, reducing it to a single-dimensional metric, which fails to capture the hierarchical structure and interdependence of spatial abilities. To address this gap, we propose a hierarchical spatial cognition framework that decomposes spatial intelligence into five progressively complex levels from basic observation to high-level planning. Building upon this taxonomy, we construct SpatialBench, a large-scale, fine-grained benchmark covering 15 tasks aligned with these cognitive levels. To provide a unified evaluation across heterogeneous tasks, we further introduce a high-level capability-oriented metric that reliably assesses a model's overall spatial reasoning ability. Extensive experiments over massive MLLMs reveal distinct performance stratification across cognitive levels: models exhibit strong perceptual grounding yet remain limited in symbolic reasoning, causal inference, and planning. Additional human tests demonstrate that humans perform selective, goal-directed abstraction, while MLLMs tend to over-attend to surface details without coherent spatial intent. Our work establishes the first systematic framework for measuring hierarchical spatial cognition in MLLMs, laying the foundation for future spatially intelligent systems.