Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 vs. gpt-oss-120b (high), Codex
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 against gpt-oss-120b (high), Codex: Geospatial Agent Index, each evaluation, cost, time and tokens, on the same tasks and limits. The harnesses differ, so part of any gap can come from the harness. Methodology.
Side by side
| Metric | Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 | gpt-oss-120b (high), Codex |
|---|---|---|
| Geospatial Agent Index: Text only | 11 | 43 |
| Geospatial Agent Index: Text + vision | No data | No image input |
| Earth-Bench V2 score | 16 | 49 |
| GeoBenchX score | 12 | 55 |
| GS-QA2 score | 4 | 25 |
| Vision agents score | No data | No image input |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 10 | 58 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 32 | 42 |
| OGD4All score (standalone; in no index) | No data | No data |
| GeoNatureAgent score (standalone; in no index) | No data | No data |
| MultiGlobeQA score (standalone; in no index) | No data | No data |
| GeoAgentEval score (standalone; in no index) | No data | No data |
| Remote sensing and raster analysis score | 11 | 40 |
| GIS analysis score | 7 | 23 |
| Climate and time series score | 15 | 44 |
| Spatial statistics score | 0 | 0 |
| Making maps score | 7 | 61 |
| Recognising infeasible requests score | 12 | 62 |
| Cost per task | $0.012 | $0.26 |
| Time per task (at 100 output tokens/s) | 4.9 min | 6.0 min |
| Tokens per attempt | 249k | 700k |
| Turns per attempt | 17.3 | 29.8 |
| Input price per 1M tokens | $0.070 | $0.35 |
| Output price per 1M tokens | $0.20 | $0.75 |
| Context window | 262,144 | 128,000 |
| Image input | No | No |
| Developer | NVIDIA | OpenAI |
Charts
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 | Developer | Geospatial Agent Index |
|---|---|---|
| gpt-oss-120b (high), Codex | OpenAI | 43 |
| Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 | NVIDIA | 11 |
Cost per task
Average cost per task (USD) · Lower is better
Data table: Cost per task
| Model | Developer | Cost per task |
|---|---|---|
| Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 | NVIDIA | $0.012 |
| gpt-oss-120b (high), Codex | OpenAI | $0.26 |
Output tokens per task
Average output tokens per attempt, reasoning included · Lower is better
Data table: Output tokens per task
| Model | Developer | Output tokens per task |
|---|---|---|
| Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 | NVIDIA | 6k |
| gpt-oss-120b (high), Codex | OpenAI | 27k |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 or gpt-oss-120b (high), Codex?
gpt-oss-120b (high), Codex, 43 to 11.
Which is cheaper per task, Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 or gpt-oss-120b (high), Codex?
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2, $0.012 to $0.26.
Which writes fewer output tokens per task, Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 or gpt-oss-120b (high), Codex?
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2, 6k to 27k.
Which has the larger context window, Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 or gpt-oss-120b (high), Codex?
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2, 262,144 tokens to 128,000 tokens.
More: Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 · gpt-oss-120b (high), Codex · All comparisons