Routing benchmark for geospatial AI
How well AI models do routing on satellite, aerial and map data: 3 sets with results, scored the same way for every model.
Sets
| Set | Index | Tasks of this type | Best score on the set | Models |
|---|---|---|---|---|
| GS-QA2 | Spatial Agent Index | 30 | 94 | 48 |
| GS-QA2 Full | Spatial Agent Index | 554 | 88 | 9 |
| OpenEarthAgent Maps | Spatial Agent Index | 83 | 100 | 14 |
Top models
Mean of each model's best score on the 3 sets, 0 to 100, for the 8 models that have run all of them.
| Rank | Model | Developer | Score |
|---|---|---|---|
| 1 | Claude Opus 5.5 | Anthropic | 94 |
| 2 | Claude Opus 5 | Anthropic | 92 |
| 3 | Claude Sonnet 5.5 | Anthropic | 92 |
| 4 | Claude Haiku 5.5 | Anthropic | 90 |
| 5 | Claude Opus 4.8 | Anthropic | 87 |
| 6 | Claude Opus 4.7 | Anthropic | 84 |
| 7 | Claude Sonnet 5 | Anthropic | 83 |
| 8 | Claude Haiku 4.5 | Anthropic | 75 |
Every task type · Every evaluation · Methodology
Frequently asked questions
Which AI model is best at routing on geospatial data?
Claude Opus 5.5 leads, on the sets below. Mean of each model's best score on the 3 sets, 0 to 100, for the 8 models that have run all of them.
What is this routing benchmark?
Every task in our sets that asks for routing, in GIS and Earth observation work, with each set's own score. It is a view over the sets, not an index of its own.
Which index does it count in?
Each set counts in one index: the Spatial Agent Index, as the table shows.
