OVO-S-Bench - Allocentric Spatial Mapping: leaderboard
Metric: Level 4 (allocentric spatial mapping: global direction, topology, trajectory-map alignment) multiple-choice accuracy (%) on the 1,680-question evaluation set, prefix-access mode (128 frames sampled uniformly from the video before the query timestamp, question given with the frames), mean of its three task families. Source: arxiv.org. Saturation forecast: Around 2032. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 54.9 |
| 2 | Qwen 3.5 27B (Non-reasoning) | 47.72 |
| 3 | Qwen 3 VL 235B A22B | 45.71 |
| 4 | Qwen 3.5 397B A17B | 45.36 |
| 5 | Qwen 2.5 VL 7B Instruct | 44.7 |
| 6 | Gemini 3.1 Flash Lite | 42.84 |
| 7 | Qwen 3 VL 4B Instruct | 41.35 |
| 8 | Qwen 3 VL 32B Instruct | 41.22 |
| 9 | GPT-5.4 | 40.53 |
| 10 | InternVL3.5-8B | 39.32 |
| 11 | Qwen 3.5 4B (Non-reasoning) | 38.73 |
| 12 | Qwen 3.5 9B (Non-reasoning) | 36.61 |
| 13 | Grok 4.1 Fast (Non-reasoning) | 35.04 |
| 14 | Gemma 4 E4B | 32.26 |
| 15 | Gemma 4 E2B | 29.62 |
Interactive version: theaggregate.ai/benchmark?slug=ovo-s-bench-allocentric-spatial-mapping · How It Works · Data refreshed daily, snapshot 2026-09-29.