Axis Spatial

Terminus-2 – Gemini 3.8 Flash (low) vs. Claude Code – Claude Haiku 5.5 (medium): geospatial agent comparison

Comparison between Terminus-2 – Gemini 3.8 Flash (low) and Claude Code – Claude Haiku 5.5 (medium) 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 – Gemini 3.8 Flash (low) vs. Claude Code – Claude Haiku 5.5 (medium)
MetricTerminus-2 – Gemini 3.8 Flash (low)Claude Code – Claude Haiku 5.5 (medium)
Geospatial Agent Index64* (attempts per task: 2.9)65* (attempts per task: 4)
GeoAgentBench score9694
GeoBenchX score6155
Earth-Bench score3718
GeoAnalystBench score6395
Raster and remote sensing score6247
Vector and overlay score6378
Networks and routing score100100
Climate and time series score5129
Spatial statistics and interpolation score4335
Mapping and cartography score5371
Recognising infeasible requests score8356
Cost per task$0.115$0.0074
Time per task2.7 min0.8 min
Tokens per attempt314k237k
Turns per attempt17.96.5
Attempts decided833745
Input price per 1M tokens$0.75$0.10
Output price per 1M tokens$3.75$0.50
Context window1,048,576No data
Image inputYesNo data
DeveloperGoogleAnthropic

Charts

Geospatial Agent Index

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

  • Partial coverage
Geospatial Agent IndexClaude Code – Claude Haiku 5.5 (medium): 65* (partial coverage); Terminus-2 – Gemini 3.8 Flash (low): 64* (partial coverage)025507510065*Claude Haiku 5.5 (medium)64*Gemini 3.8 Flash (low)
Data table: Geospatial Agent Index
Geospatial Agent Index
ModelCreatorGeospatial Agent IndexRange while attempts are pendingBenchmarks coveredCoverage (attempts)
Claude Code – Claude Haiku 5.5 (medium)Anthropic65* (attempts per task: 4)47 to 804 of 4745 of 1,167 planned attempts
Terminus-2 – Gemini 3.8 Flash (low)Google64* (attempts per task: 2.9)No pending judgements; coverage incomplete4 of 4833 of 1,167 planned attempts (74 excluded)

Cost per task

Average cost per task (USD) · Lower is better

  • Partial coverage
Cost per taskClaude Code – Claude Haiku 5.5 (medium): $0.0074; Terminus-2 – Gemini 3.8 Flash (low): $0.115 (partial coverage)$0$0.030$0.060$0.090$0.120$0.0074Claude Haiku 5.5 (medium)$0.115Gemini 3.8 Flash (low)
Data table: Cost per task
Cost per task
ModelCreatorCost per taskCoverage (attempts)
Terminus-2 – Gemini 3.8 Flash (low)Google$0.115833 of 1,167 planned attempts (74 excluded)
Claude Code – Claude Haiku 5.5 (medium)Anthropic$0.00741,167 of 1,167 planned attempts

Time per task

Average agent wall time per task · Lower is better

  • Partial coverage
Time per taskClaude Code – Claude Haiku 5.5 (medium): 0.8 min; Terminus-2 – Gemini 3.8 Flash (low): 2.7 min (partial coverage)0m1m2m2m3m0.8 minClaude Haiku 5.5 (medium)2.7 minGemini 3.8 Flash (low)
Data table: Time per task
Time per task
ModelCreatorTime per taskCoverage (attempts)
Terminus-2 – Gemini 3.8 Flash (low)Google2.7 min833 of 1,167 planned attempts (74 excluded)
Claude Code – Claude Haiku 5.5 (medium)Anthropic0.8 min1,167 of 1,167 planned attempts

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Terminus-2 – Gemini 3.8 Flash (low) or Claude Code – Claude Haiku 5.5 (medium)?
Claude Code – Claude Haiku 5.5 (medium), at 65 against 64 for Terminus-2 – Gemini 3.8 Flash (low). Terminus-2 – Gemini 3.8 Flash (low): 833 of 1,167 planned attempts (74 excluded). Claude Code – Claude Haiku 5.5 (medium): 745 of 1,167 planned attempts. Coverage is partial; scores can change as more attempts are decided.
Which is cheaper per task, Terminus-2 – Gemini 3.8 Flash (low) or Claude Code – Claude Haiku 5.5 (medium)?
Claude Code – Claude Haiku 5.5 (medium), at $0.0074 against $0.115 for Terminus-2 – Gemini 3.8 Flash (low).
Which is faster per task, Terminus-2 – Gemini 3.8 Flash (low) or Claude Code – Claude Haiku 5.5 (medium)?
Claude Code – Claude Haiku 5.5 (medium), at 0.8 min against 2.7 min for Terminus-2 – Gemini 3.8 Flash (low).
Which has the larger context window, Terminus-2 – Gemini 3.8 Flash (low) or Claude Code – Claude Haiku 5.5 (medium)?
Not known yet: Claude Code – Claude Haiku 5.5 (medium) has no context window data.

More: Terminus-2 – Gemini 3.8 Flash (low) · Claude Code – Claude Haiku 5.5 (medium) · all comparisons