Axis Spatial

Claude Code – Claude Haiku 5.5 (high) vs. Claude Code – Claude Sonnet 5.5 (xhigh): geospatial agent comparison

Comparison between Claude Code – Claude Haiku 5.5 (high) and Claude Code – Claude Sonnet 5.5 (xhigh) on the Geospatial Agent Index, each benchmark, cost, time and token use, from the same tasks, agent and limits. Methodology.

Side by side

Claude Code – Claude Haiku 5.5 (high) vs. Claude Code – Claude Sonnet 5.5 (xhigh)
MetricClaude Code – Claude Haiku 5.5 (high)Claude Code – Claude Sonnet 5.5 (xhigh)
Geospatial Agent Index75* (partial: 1.7 of 3 attempts)74 (partial: 1 of 3 attempts)
GeoAgentBench score9894
GeoBenchX score6165
Earth-Bench score5148
GeoAnalystBench score8989
Raster and remote sensing score6971
Vector and overlay score7567
Networks and routing score100100
Climate and time series score7971
Spatial statistics and interpolation score5858
Mapping and cartography score7575
Recognising infeasible requests score5663
Cost per task$0.010$0.163
Time per task1.1 min1.4 min
Tokens per attempt328k217k
Turns per attempt8.17.9
Attempts decided422290
Input price per 1M tokens$0.10$2.00
Output price per 1M tokens$0.50$10.00
Context windowNo dataNo data
Image inputNo dataNo data
DeveloperAnthropicAnthropic

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 (high): 75* (partial coverage); Claude Code – Claude Sonnet 5.5 (xhigh): 74025507510075*Claude Haiku 5.5 (high)74Claude Sonnet 5.5 (xhigh)
Data table: Geospatial Agent Index
Geospatial Agent Index
ModelCreatorGeospatial Agent IndexRange while attempts are pendingBenchmarks coveredCoverage (attempts)
Claude Code – Claude Haiku 5.5 (high)Anthropic75* (partial: 1.7 of 3 attempts)66 to 774 of 4422 of 870 planned attempts
Claude Code – Claude Sonnet 5.5 (xhigh)Anthropic74 (partial: 1 of 3 attempts)No pending judgements4 of 4290 of 290 planned attempts

Cost per task

Average cost per task (USD) · Lower is better

  • Partial coverage
Cost per taskClaude Code – Claude Haiku 5.5 (high): $0.010 (partial coverage); Claude Code – Claude Sonnet 5.5 (xhigh): $0.163$0$0.050$0.100$0.150$0.200$0.010Claude Haiku 5.5 (high)$0.163Claude Sonnet 5.5 (xhigh)
Data table: Cost per task
Cost per task
ModelCreatorCost per taskCoverage (attempts)
Claude Code – Claude Haiku 5.5 (high)Anthropic$0.010485 of 870 planned attempts
Claude Code – Claude Sonnet 5.5 (xhigh)Anthropic$0.163290 of 290 planned attempts

Time per task

Average agent wall time per task · Lower is better

  • Partial coverage
Time per taskClaude Code – Claude Haiku 5.5 (high): 1.1 min (partial coverage); Claude Code – Claude Sonnet 5.5 (xhigh): 1.4 min0m0m1m1m2m1.1 minClaude Haiku 5.5 (high)1.4 minClaude Sonnet 5.5 (xhigh)
Data table: Time per task
Time per task
ModelCreatorTime per taskCoverage (attempts)
Claude Code – Claude Haiku 5.5 (high)Anthropic1.1 min485 of 870 planned attempts
Claude Code – Claude Sonnet 5.5 (xhigh)Anthropic1.4 min290 of 290 planned attempts

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Claude Code – Claude Haiku 5.5 (high) or Claude Code – Claude Sonnet 5.5 (xhigh)?
Claude Code – Claude Haiku 5.5 (high), at 75 against 74 for Claude Code – Claude Sonnet 5.5 (xhigh). Claude Code – Claude Haiku 5.5 (high): 422 of 870 planned attempts. Claude Code – Claude Sonnet 5.5 (xhigh): 290 of 290 planned attempts. Coverage is partial; scores can change as more attempts are decided.
Which is cheaper per task, Claude Code – Claude Haiku 5.5 (high) or Claude Code – Claude Sonnet 5.5 (xhigh)?
Claude Code – Claude Haiku 5.5 (high), at $0.010 against $0.163 for Claude Code – Claude Sonnet 5.5 (xhigh).
Which is faster per task, Claude Code – Claude Haiku 5.5 (high) or Claude Code – Claude Sonnet 5.5 (xhigh)?
Claude Code – Claude Haiku 5.5 (high), at 1.1 min against 1.4 min for Claude Code – Claude Sonnet 5.5 (xhigh).
Which has the larger context window, Claude Code – Claude Haiku 5.5 (high) or Claude Code – Claude Sonnet 5.5 (xhigh)?
Not known yet: Claude Code – Claude Haiku 5.5 (high) has no context window data and Claude Code – Claude Sonnet 5.5 (xhigh) has no context window data.

More: Claude Code – Claude Haiku 5.5 (high) · Claude Code – Claude Sonnet 5.5 (xhigh) · all comparisons