Claude Sonnet 4.6 (high) vs Nemotron 3.5 Super 120B (Reasoning)
Results updated 9 November 2026 · Both in Terminus-2 · How we benchmark
Verdict
Not comparable yet: Claude Sonnet 4.6 (high) is not yet ranked.
- Cost per task
- $0.25 vs $0.011
- Time per task
- 3.2 min vs 13.8 min
- Tokens per attempt
- 76k vs 105k
- Image input
- Yes vs No
Claude Sonnet 4.6 (high), Terminus-2 scores higher than Nemotron 3.5 Super 120B (Reasoning), Terminus-2 on all 2 evaluations scored for both. Nemotron 3.5 Super 120B (Reasoning), Terminus-2 is cheaper ($0.011 against $0.25 per task), but Claude Sonnet 4.6 (high), Terminus-2 is faster (3.2 min against 13.8 min).
Not yet ranked. Claude Sonnet 4.6 (high), Terminus-2: no GS-QA2 score yet. Not in the Agent Index, cost and time charts until it has a score on every index evaluation. Results so far are in the side-by-side table.
Where they differ most
By kind of work, share of tasks passed, 0 to 100 · Higher is better
| Kind of work | Claude Sonnet 4.6 (high) | Nemotron 3.5 Super 120B (Reasoning) |
|---|---|---|
| Change detection | 25 | 56 |
| GIS analysis | 40 | 70 |
| Climate and time series | 73 | 50 |
| Remote sensing and raster analysis | 63 | 54 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Claude Sonnet 4.6 (high), Terminus-2 · Nemotron 3.5 Super 120B (Reasoning), Terminus-2.
Side by side
| Metric | Claude Sonnet 4.6 (high), Terminus-2 | Nemotron 3.5 Super 120B (Reasoning), Terminus-2 |
|---|---|---|
| Geospatial Agent Index: Text only | 58 | 56 |
| Geospatial Agent Index: Text + vision | No data | No data |
| Earth-Bench V2 score | 59 | 49 |
| GeoBenchX score | 57 | 56 |
| GS-QA2 score | No data | 63 |
| Vision agents score | No data | No data |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 96 | 83 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 100 | 79 |
| OGD4All score (standalone; in no index) | No data | 85 |
| GeoNatureAgent score (standalone; in no index) | No data | 100 |
| MultiGlobeQA score (standalone; in no index) | No data | 70 |
| GeoAgentEval score (standalone; in no index) | No data | 84 |
| ImpactMesh score (standalone; in no index) | No data | No data |
| Remote sensing and raster analysis score | 63 | 54 |
| Change detection score | 25 | 56 |
| GIS analysis score | 40 | 70 |
| Climate and time series score | 73 | 50 |
| Making maps score | 75 | 70 |
| Cost per task | $0.25 | $0.011 |
| Time per task (at 100 output tokens/s) | 3.2 min | 13.8 min |
| Tokens per attempt | 76k | 105k |
| Turns per attempt | 6.9 | 10.5 |
| Input price per 1M tokens | $3.00 | $0.20 |
| Output price per 1M tokens | $15.00 | $0.50 |
| Context window | 1,000,000 | 262,144 |
| Image input | Yes | No |
| Developer | Anthropic | NVIDIA |