Axis Spatial

Terminus-2 – DeepSeek V4 Flash (high) vs. Codex – GPT-6 Luna (xhigh): geospatial agent comparison

Comparison between Terminus-2 – DeepSeek V4 Flash (high) and Codex – GPT-6 Luna (xhigh) on the Geospatial Agent Index, each benchmark, cost, time and token use, from the same tasks, agent and limits. Methodology.

Side by side

Terminus-2 – DeepSeek V4 Flash (high) vs. Codex – GPT-6 Luna (xhigh)
MetricTerminus-2 – DeepSeek V4 Flash (high)Codex – GPT-6 Luna (xhigh)
Geospatial Agent Index70*77* (partial: 1.9 of 3 attempts)
GeoAgentBench score9599
GeoBenchX score5460
Earth-Bench score3755
GeoAnalystBench score9395
Raster and remote sensing score6675
Vector and overlay score7278
Networks and routing score87100
Climate and time series score5875
Spatial statistics and interpolation score5863
Mapping and cartography score7664
Recognising infeasible requests score4062
Cost per task$0.067$0.0092
Time per task9.7 min3.4 min
Tokens per attempt312k236k
Turns per attempt14.39.8
Attempts decided860512
Input price per 1M tokens$0.44$0.10
Output price per 1M tokens$1.32$0.50
Context window1,048,576No data
Image inputNoNo data
DeveloperDeepSeekOpenAI

Charts

Geospatial Agent Index

Equal-weight average of the evaluation scores, 0 to 100 · Higher is better

  • Partial coverage
Geospatial Agent IndexCodex – GPT-6 Luna (xhigh): 77* (partial coverage); Terminus-2 – DeepSeek V4 Flash (high): 70* (partial coverage)025507510077*GPT-6 Luna (xhigh)70*DeepSeek V4 Flash (high)
Data table: Geospatial Agent Index
Geospatial Agent Index
ModelCreatorGeospatial Agent IndexRange while attempts are pendingBenchmarks coveredCoverage (attempts)
Codex – GPT-6 Luna (xhigh)OpenAI77* (partial: 1.9 of 3 attempts)72 to 784 of 4512 of 870 planned attempts
Terminus-2 – DeepSeek V4 Flash (high)DeepSeek70*70 to 714 of 4860 of 870 planned attempts

Cost per task

Average cost per task (USD) · Lower is better

  • Partial coverage
Cost per taskCodex – GPT-6 Luna (xhigh): $0.0092 (partial coverage); Terminus-2 – DeepSeek V4 Flash (high): $0.067 (partial coverage)$0$0.020$0.040$0.060$0.080$0.0092GPT-6 Luna (xhigh)$0.067DeepSeek V4 Flash (high)
Data table: Cost per task
Cost per task
ModelCreatorCost per taskCoverage (attempts)
Terminus-2 – DeepSeek V4 Flash (high)DeepSeek$0.067869 of 870 planned attempts
Codex – GPT-6 Luna (xhigh)OpenAI$0.0092560 of 870 planned attempts

Time per task

Average agent wall time per task · Lower is better

  • Partial coverage
Time per taskCodex – GPT-6 Luna (xhigh): 3.4 min (partial coverage); Terminus-2 – DeepSeek V4 Flash (high): 9.7 min (partial coverage)0m2m5m8m10m3.4 minGPT-6 Luna (xhigh)9.7 minDeepSeek V4 Flash (high)
Data table: Time per task
Time per task
ModelCreatorTime per taskCoverage (attempts)
Terminus-2 – DeepSeek V4 Flash (high)DeepSeek9.7 min869 of 870 planned attempts
Codex – GPT-6 Luna (xhigh)OpenAI3.4 min560 of 870 planned attempts

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Terminus-2 – DeepSeek V4 Flash (high) or Codex – GPT-6 Luna (xhigh)?
Codex – GPT-6 Luna (xhigh), at 77 against 70 for Terminus-2 – DeepSeek V4 Flash (high). Terminus-2 – DeepSeek V4 Flash (high): 860 of 870 planned attempts. Codex – GPT-6 Luna (xhigh): 512 of 870 planned attempts. Coverage is partial; scores can change as more attempts are decided.
Which is cheaper per task, Terminus-2 – DeepSeek V4 Flash (high) or Codex – GPT-6 Luna (xhigh)?
Codex – GPT-6 Luna (xhigh), at $0.0092 against $0.067 for Terminus-2 – DeepSeek V4 Flash (high).
Which is faster per task, Terminus-2 – DeepSeek V4 Flash (high) or Codex – GPT-6 Luna (xhigh)?
Codex – GPT-6 Luna (xhigh), at 3.4 min against 9.7 min for Terminus-2 – DeepSeek V4 Flash (high).
Which has the larger context window, Terminus-2 – DeepSeek V4 Flash (high) or Codex – GPT-6 Luna (xhigh)?
Not known yet: Codex – GPT-6 Luna (xhigh) has no context window data.

More: Terminus-2 – DeepSeek V4 Flash (high) · Codex – GPT-6 Luna (xhigh) · all comparisons