StepFun models: geospatial agent results
Geospatial Agent Benchmarks tests 1 entry from StepFun.
Highlights
- Highest Geospatial Agent Index (ranked entries): Step 5 Preview (Reasoning), Terminus-2 (34).
- Lowest cost per task (ranked entries): Step 5 Preview (Reasoning), Terminus-2 ($0.071).
- Fewest output tokens per task (ranked entries): Step 5 Preview (Reasoning), Terminus-2 (10k).
Results
Geospatial Agent Index
Equal average of Earth-Bench V2, GeoBenchX and GS-QA2, 0 to 100 · Higher is better
Data table: Geospatial Agent Index
| Model | Geospatial Agent Index: Text only | Earth-Bench V2 | GeoBenchX | GS-QA2 | Vision agents | Cost per task | Time per task (at 100 output tokens/s) | Context window |
|---|---|---|---|---|---|---|---|---|
| Step 5 Preview (Reasoning), Terminus-2 | 34 | 32 | 17 | 52 | No data | $0.071 | 7.9 min (estimated) | 1,024,000 |