OmniCoT (Direct Answer): leaderboard
OmniCoT (2026), a panoramic spatial reasoning benchmark, answered directly without chain of thought: overall accuracy in percent on the synthetic OmniCoT-B panoramas. Higher is better.
Metric: Overall accuracy (%) when answering directly without chain of thought on OmniCoT-B (synthetic 360-degree equirectangular panoramas with three anchor objects given by coordinates and orientation); answers are object identities, visibility, directions or the closest angle option, exact match after normalization. Source: arxiv.org. Saturation forecast: Around June 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 42.66 | #167 |
| 2 | Gemini 3 Flash | 39.97 | #164 |
| 3 | Qwen 2.5 VL 72B Instruct | 26.51 | #846 |
| 4 | Qwen 3 VL 8B Instruct | 25.39 | #981 |
| 5 | Qwen 2.5 VL 7B Instruct | 23.01 | #1833 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=omnicot-direct-answer · How It Works · Data refreshed daily, snapshot 2026-11-13.