SpatiO: Adaptive Test-Time Orchestration of
Vision-Language Agents for Spatial Reasoning

Understanding visual scenes requires not only recognizing objects but also reasoning about their spatial relationships. Unlike general vision-language tasks, spatial reasoning requires integrating multiple inductive biases—such as 2D appearance cues, depth signals, and geometric constraints—whose reliability varies across contexts. This suggests that effective spatial reasoning requires spatial adaptability: the ability to flexibly coordinate different reasoning strategies depending on the input.

However, most existing approaches rely on a single reasoning pipeline that implicitly learns a fixed spatial prior, limiting their ability to adapt under distribution changes. Multi-agent systems offer a promising alternative by aggregating diverse reasoning trajectories, but prior attempts in spatial reasoning primarily employ homogeneous agents, restricting the diversity of inductive biases they can leverage.

In this work, we introduce SpatiO, a heterogeneous multi-agent framework for spatial reasoning. SpatiO adaptively orchestrates specialists with diverse spatial inductive biases at test time, guided by continuously updated confidence scores that reflect each specialist's past reliability on similar queries.