OmniMapBench - Single-Step Reasoning: leaderboard
Level 2 of OmniMapBench (2026): strict exact-match accuracy in percent on the 802 single-step reasoning questions about map documents, with chain-of-thought prompting. Higher is better.
Metric: Level 2 (single-step reasoning, 802 questions): strict exact-match accuracy on manually annotated single-choice, multiple-choice and ordering questions over map documents, chain-of-thought prompting (%); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 81.53 | #94 |
| 2 | Qwen 3.5 397B A17B | 79.01 | #246 |
| 3 | Qwen 3.5 Plus | 77.64 | #232 |
| 4 | Qwen 3.5 27B | 77.1 | #359 |
| 5 | Qwen 3.5 122B A10B | 76.52 | #331 |
| 6 | Kimi K2.5 | 74.8 | #256 |
| 7 | Qwen 3.5 Flash | 73.87 | #435 |
| 8 | Qwen 3.5 35B A3B | 72.27 | #482 |
| 9 | Gemini 2.5 Pro | 70.7 | #266 |
| 10 | GPT-5 Mini | 62.59 | #348 |
| 11 | GPT-5 | 61.47 | #167 |
| 12 | Gemma 4 31B (IT) | 58.35 | #342 |
| 13 | GLM-4.5V | 54.24 | #744 |
| 14 | Claude Sonnet 4.5 | 52.63 | #250 |
| 15 | Gemma 4 26B A4B (IT) | 49.75 | #549 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=omnimapbench-single-step-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-13.