GeoBench ACW — leaderboard
GeoGuessr for LLMs (ACW subset) - identify location from a single static image using the Around-the-Clock-and-World dataset. Tests visual geographic reasoning and world knowledge.
Metric: Average Score. Source: geobench.org. Status: saturation imminent. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5 | 4403.7 |
| 2 | Gemini 3 Flash (Preview) | 4332.77 |
| 3 | Gemini 3.5 Flash | 4288.21 |
| 4 | Gemini 3.1 Flash Lite (Preview) | 4229.08 |
| 5 | Gemini 3.1 Pro (Preview) | 4172.19 |
| 6 | Gemini 3 Pro (Preview) | 4144.58 |
| 7 | Claude Opus 4.8 (Thinking) | 4127.75 |
| 8 | Human Expert | 4100 |
| 9 | Gemini 2.5 Pro (Preview 03-25) | 4092.5 |
| 10 | Gemini 2.5 Pro | 4019.35 |
| 11 | GPT-5 | 3937 |
| 12 | Gemini 1.5 Flash | 3933.94 |
| 13 | O3 (High) | 3929.63 |
| 14 | O1 | 3924.6 |
| 15 | Gemini 2.0 Flash | 3897.45 |
Interactive version: theaggregate.ai/benchmark?slug=geobench-acw · How the rankings work · Data refreshed daily, snapshot 2026-07-22.