RoboStressBench - Spatial Reasoning: leaderboard

Metric: Accuracy (%) on the 1,369 spatial-reasoning multiple-choice questions, exact match on the predicted option, greedy decoding with at most 64 new tokens (API models: temperature 0, 64 output tokens), one unified instruction template; images of embodied scenes under physical visual stress (material, viewpoint, lighting, geometry); higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 16 models tracked.

Top models

#ModelScore
1GPT-5.580.3
2Qwen 3.5 35B A3B79
3Qwen 3.6 27B78.3
4Qwen 3.6 35B A3B78.1
5Qwen 3.5 27B77.1
6Qwen 3.5 9B73.7
7Qwen 3.5 4B72.6
8Gemini 3.1 Pro (Preview)70
9InternVL3.5-8B67.9
10Molmo2-8B63.6

Interactive version: theaggregate.ai/benchmark?slug=robostressbench-spatial-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-29.