Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) vs. Codex – GPT-6.1 Sol (high): geospatial agent comparison
Comparison between Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) and Codex – GPT-6.1 Sol (high) on the Geospatial Agent Index, each benchmark, cost, time and token use, from the same tasks, agent and limits. Methodology.
Side by side
| Metric | Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) | Codex – GPT-6.1 Sol (high) |
|---|---|---|
| Geospatial Agent Index | 74* (partial: 1.1 of 3 attempts) | 80* (partial: 2.8 of 3 attempts) |
| GeoAgentBench score | 97 | 100 |
| GeoBenchX score | 51 | 66 |
| Earth-Bench score | 54 | 53 |
| GeoAnalystBench score | 95 | 100 |
| Raster and remote sensing score | 72 | 72 |
| Vector and overlay score | 70 | 78 |
| Networks and routing score | 100 | 100 |
| Climate and time series score | 83 | 76 |
| Spatial statistics and interpolation score | 71 | 61 |
| Mapping and cartography score | 75 | 76 |
| Recognising infeasible requests score | 30 | 68 |
| Cost per task | $0.278 | $0.080 |
| Time per task | 3.7 min | 1.8 min |
| Tokens per attempt | 71k | 143k |
| Turns per attempt | 6.3 | 7.8 |
| Attempts decided | 306 | 753 |
| Input price per 1M tokens | $3.00 | $2.00 |
| Output price per 1M tokens | $15.00 | $10.00 |
| Context window | 1,000,000 | No data |
| Image input | Yes | No data |
| Developer | Anthropic | OpenAI |
Charts
Geospatial Agent Index
Equal-weight average of the evaluation scores, 0 to 100 · Higher is better
First eight selected entries in metric order. Full results in table view.
Data table: Geospatial Agent Index
| Model | Creator | Geospatial Agent Index | Range while attempts are pending | Benchmarks covered | Coverage (attempts) |
|---|---|---|---|---|---|
| Codex – GPT-6.1 Sol (high) | OpenAI | 80* (partial: 2.8 of 3 attempts) | 72 to 80 | 4 of 4 | 753 of 870 planned attempts |
| Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) | Anthropic | 74* (partial: 1.1 of 3 attempts) | 73 to 75 | 4 of 4 | 306 of 870 planned attempts (133 excluded) |
Codex – GPT-6.1 Sol (high)
Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high)Cost per task
Average cost per task (USD) · Lower is better
First eight selected entries in metric order. Full results in table view.
Data table: Cost per task
| Model | Creator | Cost per task | Coverage (attempts) |
|---|---|---|---|
| Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) | Anthropic | $0.278 | 321 of 870 planned attempts (133 excluded) |
| Codex – GPT-6.1 Sol (high) | OpenAI | $0.080 | 813 of 870 planned attempts |
Time per task
Average agent wall time per task · Lower is better
First eight selected entries in metric order. Full results in table view.
Data table: Time per task
| Model | Creator | Time per task | Coverage (attempts) |
|---|---|---|---|
| Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) | Anthropic | 3.7 min | 321 of 870 planned attempts (133 excluded) |
| Codex – GPT-6.1 Sol (high) | OpenAI | 1.8 min | 813 of 870 planned attempts |
Frequently asked questions
- Which scores higher on the Geospatial Agent Index, Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) or Codex – GPT-6.1 Sol (high)?
- Codex – GPT-6.1 Sol (high), at 80 against 74 for Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high). Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high): 306 of 870 planned attempts (133 excluded). Codex – GPT-6.1 Sol (high): 753 of 870 planned attempts. Coverage is partial; scores can change as more attempts are decided.
- Which is cheaper per task, Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) or Codex – GPT-6.1 Sol (high)?
- Codex – GPT-6.1 Sol (high), at $0.080 against $0.278 for Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high).
- Which is faster per task, Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) or Codex – GPT-6.1 Sol (high)?
- Codex – GPT-6.1 Sol (high), at 1.8 min against 3.7 min for Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high).
- Which has the larger context window, Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) or Codex – GPT-6.1 Sol (high)?
- Not known yet: Codex – GPT-6.1 Sol (high) has no context window data.
More: Terminus-2 – Claude Sonnet 4.6 (adaptive, effort high) · Codex – GPT-6.1 Sol (high) · all comparisons