Spatial reasoning is a fundamental aspect of human cognition, yet it remains a major challenge for contemporary vision-language models (VLMs). Prior work largely relied on synthetic or LLM-generated environments with limited task designs and puzzle-like setups, failing to capture the real-world complexity, visual noise, and diverse spatial relationships that VLMs encounter. To address this, we introduce SpatiaLab, a comprehensive benchmark for evaluating VLMs' spatial reasoning in realistic, unconstrained contexts. SpatiaLab comprises 1,400 visual question-answer pairs across six major categories: Relative Positioning, Depth & Occlusion, Orientation, Size & Scale, Spatial Navigation, and 3D Geometry, each with five subcategories, yielding 30 distinct task types. Each subcategory contains at least 25 questions, and each main category includes at least 200 questions, supporting both multiple-choice and open-ended evaluation. Experiments across diverse state-of-the-art VLMs, including open- and closed-source models, reasoning-focused, and specialized spatial reasoning models, reveal a substantial gap in spatial reasoning capabilities compared with humans. In the multiple-choice setup, InternVL3.5-72B achieves 54.93% accuracy versus 87.57% for humans. In the open-ended setting, all models show a performance drop of around 10-25%, with GPT-5-mini scoring highest at 40.93% versus 64.93% for humans. These results highlight key limitations in handling complex spatial relationships, depth perception, navigation, and 3D geometry. By providing a diverse, real-world evaluation framework, SpatiaLab exposes critical challenges and opportunities for advancing VLMs' spatial reasoning, offering a benchmark to guide future research toward robust, human-aligned spatial understanding. SpatiaLab is available at: https://spatialab-reasoning.github.io/.

翻译：空间推理是人类认知的基本组成部分，但仍是当代视觉语言模型面临的主要挑战。以往工作主要依赖合成环境或大语言模型生成的有限任务设计与谜题式设置，未能捕捉视觉语言模型在真实世界面临的复杂性、视觉噪声及多样化空间关系。为解决这一问题，我们提出SpatiaLab——一个在真实无约束场景中评估视觉语言模型空间推理能力的综合基准。该基准包含1400个视觉问答对，涵盖六大主要类别：相对定位、深度与遮挡、方向、尺寸与比例、空间导航及3D几何，每个类别下设五个子类别，形成30种不同任务类型。每个子类别至少包含25个问题，每个主类别至少包含200个问题，支持多选题与开放式评估。针对多种先进视觉语言模型（包括开源与闭源模型、推理专注型及专用空间推理模型）的实验表明，其空间推理能力与人类相比存在显著差距。在多选题设置中，InternVL3.5-72B准确率为54.93%，而人类达87.57%；在开放式设置中，所有模型性能下降约10-25%，最高分的GPT-5-mini仅达40.93%，人类为64.93%。这些结果凸显了模型在处理复杂空间关系、深度感知、导航及3D几何方面的关键局限。通过提供多样化真实世界评估框架，SpatiaLab揭示了推动视觉语言模型空间推理的关键挑战与机遇，为未来研究实现鲁棒的人类对齐空间理解提供基准。SpatiaLab访问地址：https://spatialab-reasoning.github.io/。