GeoBenchLLM - GeoQuestions1089 Coordinates: leaderboard
Metric: Coordinates accuracy (%). Source: arxiv.org. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-OSS-120B (Low) | 72 |
| 2 | GPT-OSS-20B (Low) | 48 |
| 3 | Qwen 3 8B (Thinking) | 43 |
| 4 | Qwen 3 8B (Non-reasoning) | 39 |
| 5 | Qwen 3 1.7B (Thinking) | 11 |
| 6 | Qwen 3 0.6B (Thinking) | 2 |
Interactive version: theaggregate.ai/benchmark?slug=geobenchllm-geoquestions1089-coordinates · How It Works · Data refreshed daily, snapshot 2026-09-19.