Terminus-2 – GLM-5.3 (low) vs. Claude Code – Claude Sonnet 5 (high): geospatial agent comparison

Comparison between Terminus-2 – GLM-5.3 (low) and Claude Code – Claude Sonnet 5 (high) 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 – GLM-5.3 (low) vs. Claude Code – Claude Sonnet 5 (high)
Metric	Terminus-2 – GLM-5.3 (low)	Claude Code – Claude Sonnet 5 (high)
Geospatial Agent Index	38*	56* (partial: attempt 1 on 104 of 320 tasks)
Earth-Bench V2 score	33	56
GeoBenchX score	43	56
GeoAgentBench score (guided GIS workflow; outside the index)	90	100
GeoAnalystBench score (guided GIS workflow; outside the index)	89	83
Raster and remote sensing score	31	52
Vector and overlay score	17	0
Networks and routing score	No data	No data
Climate and time series score	38	67
Spatial statistics and interpolation score	0	50
Mapping and cartography score	63	80
Recognising infeasible requests score	38	41
Cost per task	$0.407	$0.331
Time per task	8.7 min	2.8 min
Tokens per attempt	342k	555k
Turns per attempt	13.3	13.1
Attempts decided	320	104
Input price per 1M tokens	$1.40	$2.00
Output price per 1M tokens	$4.40	$10.00
Context window	1,048,576	No data
Image input	No	No data
Developer	Z.ai	Anthropic

Charts

Partial coverage. Claude Code – Claude Sonnet 5 (high): 41 of 147 Earth-Bench V2 tasks; 63 of 173 GeoBenchX tasks. Grey placeholders identify these entries in the index, cost and time charts; their height does not represent a value. Comparable bars appear when first-attempt coverage is complete. Their available results remain in the side-by-side table.

Geospatial Agent Index

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

Planned attempts still to come

Data table: Geospatial Agent Index

Geospatial Agent Index
Model	Creator	Geospatial Agent Index	Range while attempts are pending	Benchmarks covered	Coverage (attempts)
Terminus-2 – GLM-5.3 (low)	Z.ai	38*	No pending judgements; coverage incomplete	2 of 2	320 of 614 planned attempts

Cost per task

Average cost per task (USD) · Lower is better

Planned attempts still to come

How to read this chart

Not shown: partial coverage. Claude Code – Claude Sonnet 5 (high).

Data table: Cost per task

Cost per task
Model	Creator	Cost per task	Coverage (attempts)
Terminus-2 – GLM-5.3 (low)	Z.ai	$0.407	320 of 614 planned attempts

Time per task

Average agent wall time per task · Lower is better

Planned attempts still to come

How to read this chart

Not shown: partial coverage. Claude Code – Claude Sonnet 5 (high).

Data table: Time per task

Time per task
Model	Creator	Time per task	Coverage (attempts)
Terminus-2 – GLM-5.3 (low)	Z.ai	8.7 min	320 of 614 planned attempts

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Terminus-2 – GLM-5.3 (low) or Claude Code – Claude Sonnet 5 (high)?

Not comparable yet: Claude Code – Claude Sonnet 5 (high) has partial coverage (41 of 147 Earth-Bench V2 tasks; 63 of 173 GeoBenchX tasks), so it is not ranked. Terminus-2 – GLM-5.3 (low): 320 of 614 planned attempts. Claude Code – Claude Sonnet 5 (high): 104 of 614 planned attempts. Coverage is partial; scores can change as more attempts are decided.

Which is cheaper per task, Terminus-2 – GLM-5.3 (low) or Claude Code – Claude Sonnet 5 (high)?

Claude Code – Claude Sonnet 5 (high), at $0.331 against $0.407 for Terminus-2 – GLM-5.3 (low).

Which is faster per task, Terminus-2 – GLM-5.3 (low) or Claude Code – Claude Sonnet 5 (high)?

Claude Code – Claude Sonnet 5 (high), at 2.8 min against 8.7 min for Terminus-2 – GLM-5.3 (low).

Which has the larger context window, Terminus-2 – GLM-5.3 (low) or Claude Code – Claude Sonnet 5 (high)?

Not known yet: Claude Code – Claude Sonnet 5 (high) has no context window data.

More: Terminus-2 – GLM-5.3 (low) · Claude Code – Claude Sonnet 5 (high) · all comparisons