Nemotron 3.5 Bolt E4B (Reasoning) vs Nemotron 3.5 Super VL 120B (Reasoning)
Results updated 12 November 2026 · Both in Terminus-2 · How we benchmark
Verdict
Nemotron 3.5 Super VL 120B (Reasoning) clearly beats Nemotron 3.5 Bolt E4B (Reasoning) by 48 points, and is faster.
- Cost per task
- No data vs $0.0095
- Time per task
- 5.2 min vs 3.8 min
- Tokens per attempt
- 341k vs 98k
- Image input
- No vs Yes
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2 scores higher than Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 on all 3 evaluations scored for both.
Where they differ most
By kind of work, share of tasks passed, 0 to 100 · Higher is better
| Kind of work | Nemotron 3.5 Bolt E4B (Reasoning) | Nemotron 3.5 Super VL 120B (Reasoning) |
|---|---|---|
| Making maps | 2 | 67 |
| Remote sensing and raster analysis | 1 | 61 |
| GIS analysis | 3 | 51 |
| Climate and time series | 6 | 50 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 · Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2.
Side by side
| Metric | Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 | Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2 |
|---|---|---|
| Geospatial Agent Index: Text only | 4 | 52 |
| Geospatial Agent Index: Text + vision | No data | No data |
| Earth-Bench V2 score | 5 | 46 |
| GeoBenchX score | 3 | 53 |
| GS-QA2 score | 3 | 59 |
| Vision agents score | No data | No data |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 4 | 68 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 5 | 37 |
| OGD4All score (standalone; in no index) | 21 | 91 |
| GeoNatureAgent score (standalone; in no index) | 0 | 93 |
| MultiGlobeQA score (standalone; in no index) | 16 | 73 |
| GeoAgentEval score (standalone; in no index) | 40 | 80 |
| ImpactMesh score (standalone; in no index) | No data | No data |
| Remote sensing and raster analysis score | 1 | 61 |
| Change detection score | 8 | 44 |
| GIS analysis score | 3 | 51 |
| Climate and time series score | 6 | 50 |
| Making maps score | 2 | 67 |
| Cost per task | No data | $0.0095 |
| Time per task (at 100 output tokens/s) | 5.2 min | 3.8 min |
| Tokens per attempt | 341k | 98k |
| Turns per attempt | 24.4 | 10.1 |
| Input price per 1M tokens | No data | $0.15 |
| Output price per 1M tokens | No data | $0.50 |
| Context window | 65,536 | 262,144 |
| Image input | No | Yes |
| Developer | NVIDIA | NVIDIA |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 or Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2?
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2, 52 to 4.
Which is cheaper per task, Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 or Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2?
Not known yet: Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 has no cost data.
Which writes fewer output tokens per task, Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 or Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2?
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2, 7k to 14k.
Which has the larger context window, Nemotron 3.5 Bolt E4B (Reasoning), Terminus-2 or Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2?
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2, 262,144 tokens to 65,536 tokens.