Terminus-2 – Gemini 3.8 Flash (high) vs. Codex – GPT-6.1 Sol (xhigh): geospatial agent comparison
Comparison between Terminus-2 – Gemini 3.8 Flash (high) and Codex – GPT-6.1 Sol (xhigh) 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 – Gemini 3.8 Flash (high) | Codex – GPT-6.1 Sol (xhigh) |
|---|---|---|
| Geospatial Agent Index | 39* (attempts per task: 2.9) | 68* (attempts per task: 3.3) |
| GeoAgentBench score | 58 | 98 |
| GeoBenchX score | 40 | 59 |
| Earth-Bench score | 28 | 22 |
| GeoAnalystBench score | 32 | 95 |
| Raster and remote sensing score | 37 | 47 |
| Vector and overlay score | 44 | 78 |
| Networks and routing score | 80 | 100 |
| Climate and time series score | 26 | 41 |
| Spatial statistics and interpolation score | 17 | 42 |
| Mapping and cartography score | 25 | 74 |
| Recognising infeasible requests score | 67 | 66 |
| Cost per task | $0.308 | $0.124 |
| Time per task | 8.2 min | 3.1 min |
| Tokens per attempt | 982k | 201k |
| Turns per attempt | 32.4 | 9.0 |
| Attempts decided | 278 | 871 |
| Input price per 1M tokens | $0.75 | $2.00 |
| Output price per 1M tokens | $3.75 | $10.00 |
| Context window | 1,048,576 | No data |
| Image input | Yes | No data |
| Developer | 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 (xhigh) | OpenAI | 68* (attempts per task: 3.3) | 66 to 78 | 4 of 4 | 871 of 1,167 planned attempts |
| Terminus-2 – Gemini 3.8 Flash (high) | 39* (attempts per task: 2.9) | 13 to 80 | 4 of 4 | 278 of 1,167 planned attempts (1 excluded) |
Codex – GPT-6.1 Sol (xhigh)
Terminus-2 – Gemini 3.8 Flash (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 – Gemini 3.8 Flash (high) | $0.308 | 835 of 1,167 planned attempts (1 excluded) | |
| Codex – GPT-6.1 Sol (xhigh) | OpenAI | $0.124 | 949 of 1,167 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 – Gemini 3.8 Flash (high) | 8.2 min | 835 of 1,167 planned attempts (1 excluded) | |
| Codex – GPT-6.1 Sol (xhigh) | OpenAI | 3.1 min | 949 of 1,167 planned attempts |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Terminus-2 – Gemini 3.8 Flash (high) or Codex – GPT-6.1 Sol (xhigh)?
Codex – GPT-6.1 Sol (xhigh), at 68 against 39 for Terminus-2 – Gemini 3.8 Flash (high). Terminus-2 – Gemini 3.8 Flash (high): 278 of 1,167 planned attempts (1 excluded). Codex – GPT-6.1 Sol (xhigh): 871 of 1,167 planned attempts. Coverage is partial; scores can change as more attempts are decided.
Which is cheaper per task, Terminus-2 – Gemini 3.8 Flash (high) or Codex – GPT-6.1 Sol (xhigh)?
Codex – GPT-6.1 Sol (xhigh), at $0.124 against $0.308 for Terminus-2 – Gemini 3.8 Flash (high).
Which is faster per task, Terminus-2 – Gemini 3.8 Flash (high) or Codex – GPT-6.1 Sol (xhigh)?
Codex – GPT-6.1 Sol (xhigh), at 3.1 min against 8.2 min for Terminus-2 – Gemini 3.8 Flash (high).
Which has the larger context window, Terminus-2 – Gemini 3.8 Flash (high) or Codex – GPT-6.1 Sol (xhigh)?
Not known yet: Codex – GPT-6.1 Sol (xhigh) has no context window data.
More: Terminus-2 – Gemini 3.8 Flash (high) · Codex – GPT-6.1 Sol (xhigh) · all comparisons