Axis Spatial

DeepSeek V4 Pro (max), Terminus-2 vs. Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2

DeepSeek V4 Pro (max), Terminus-2 against Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2: Geospatial Agent Index, each evaluation, cost, time and tokens, on the same tasks and limits. Methodology.

Side by side

DeepSeek V4 Pro (max), Terminus-2 vs. Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2
MetricDeepSeek V4 Pro (max), Terminus-2Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2
Geospatial Agent Index: Text only7111
Geospatial Agent Index: Text + visionNo image inputNo data
Earth-Bench V2 score6716
GeoBenchX score6712
GS-QA2 score804
Vision agents scoreNo image inputNo data
GeoAgentBench score (Guided workflows; not in the Agent Index)9610
GeoAnalystBench score (Guided workflows; not in the Agent Index)10032
OGD4All score (standalone; in no index)No dataNo data
GeoNatureAgent score (standalone; in no index)No dataNo data
MultiGlobeQA score (standalone; in no index)No dataNo data
GeoAgentEval score (standalone; in no index)No dataNo data
Remote sensing and raster analysis score7611
GIS analysis score787
Climate and time series score6615
Spatial statistics score460
Making maps score827
Recognising infeasible requests score5512
Cost per task$0.35$0.012
Time per task (at 100 output tokens/s)4.9 min4.9 min
Tokens per attempt312k249k
Turns per attempt14.817.3
Input price per 1M tokens$1.32$0.070
Output price per 1M tokens$3.96$0.20
Context window1,048,576262,144
Image inputNoNo
DeveloperDeepSeekNVIDIA

Charts

Geospatial Agent Index

Equal average of Earth-Bench V2, GeoBenchX and GS-QA2, 0 to 100 · Higher is better

Geospatial Agent IndexScores 0 to 100: DeepSeek V4 Pro (max), Terminus-2: 71; Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2: 11025507510071DeepSeek V4 Pro (max)11Nemotron 3.5 Lightning 30B (Reasoning)
Data table: Geospatial Agent Index
Geospatial Agent Index
ModelDeveloperGeospatial Agent Index
DeepSeek V4 Pro (max), Terminus-2DeepSeek71
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2NVIDIA11

Cost per task

Average cost per task (USD) · Lower is better

Cost per taskCost per task in US dollars: Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2: $0.012; DeepSeek V4 Pro (max), Terminus-2: $0.35$0$0.10$0.20$0.30$0.40$0.012Nemotron 3.5 Lightning 30B (Reasoning)$0.35DeepSeek V4 Pro (max)
Data table: Cost per task
Cost per task
ModelDeveloperCost per task
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2NVIDIA$0.012
DeepSeek V4 Pro (max), Terminus-2DeepSeek$0.35

Output tokens per task

Average output tokens per attempt, reasoning included · Lower is better

Output tokens per taskOutput tokens per task: Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2: 6k; DeepSeek V4 Pro (max), Terminus-2: 24k010k20k30k40k6kNemotron 3.5 Lightning 30B (Reasoning)24kDeepSeek V4 Pro (max)
Data table: Output tokens per task
Output tokens per task
ModelDeveloperOutput tokens per task
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2NVIDIA6k
DeepSeek V4 Pro (max), Terminus-2DeepSeek24k

Frequently asked questions

Which scores higher on the Geospatial Agent Index, DeepSeek V4 Pro (max), Terminus-2 or Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2?
DeepSeek V4 Pro (max), Terminus-2, 71 to 11.
Which is cheaper per task, DeepSeek V4 Pro (max), Terminus-2 or Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2?
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2, $0.012 to $0.35.
Which writes fewer output tokens per task, DeepSeek V4 Pro (max), Terminus-2 or Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2?
Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2, 6k to 24k.
Which has the larger context window, DeepSeek V4 Pro (max), Terminus-2 or Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2?
DeepSeek V4 Pro (max), Terminus-2, 1,048,576 tokens to 262,144 tokens.

More: DeepSeek V4 Pro (max), Terminus-2 · Nemotron 3.5 Lightning 30B (Reasoning), Terminus-2 · All comparisons