Terminus-2 – Kimi K3 (always on) vs. Claude Code – GLM-5.2 (max): geospatial agent comparison

Comparison between Terminus-2 – Kimi K3 (always on) and Claude Code – GLM-5.2 (max) 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 – Kimi K3 (always on) vs. Claude Code – GLM-5.2 (max)
Metric	Terminus-2 – Kimi K3 (always on)	Claude Code – GLM-5.2 (max)
Geospatial Agent Index	58* (partial: attempt 1 on 320 of 565 tasks)	53* (partial: attempt 1 on 200 of 565 tasks)
Earth-Bench V2 score	61	63
GeoBenchX score	55	42
GS-QA2 score	No data	No data
GeoAgentBench score (guided GIS workflow; outside the index)	94	No data
GeoAnalystBench score (guided GIS workflow; outside the index)	89	No data
Raster and remote sensing score	59	62
Vector and overlay score	33	40
Networks and routing score	No data	No data
Climate and time series score	66	67
Spatial statistics and interpolation score	50	No data
Mapping and cartography score	78	No data
Recognising infeasible requests score	38	41
Cost per task	$0.336	$0.437
Time per task	3.9 min	7.1 min
Tokens per attempt	76k	827k
Turns per attempt	7.7	18.7
Attempts decided	320	120
Input price per 1M tokens	$3.00	$1.40
Output price per 1M tokens	$15.00	$4.40
Context window	1,000,000	262,144
Image input	Yes	No
Developer	Moonshot AI	Z.ai

Charts

Partial coverage. Terminus-2 – Kimi K3 (always on): GS-QA2 incomplete; Claude Code – GLM-5.2 (max): 35 of 147 Earth-Bench V2 tasks; 165 of 173 GeoBenchX tasks; GS-QA2 incomplete. Grey placeholders identify these entries in the index, cost and output token 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

Data table: Geospatial Agent Index

Geospatial Agent Index
Model	Creator	Geospatial Agent Index	Range while attempts are pending	Benchmarks covered	Coverage (attempts)

Cost per task

Average cost per task (USD) · Lower is better

How to read this chart

Not shown: partial coverage. Terminus-2 – Kimi K3 (always on), Claude Code – GLM-5.2 (max).

Data table: Cost per task

Cost per task
Model	Creator	Cost per task	Coverage (attempts)

Output tokens per task

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

How to read this chart

Not shown: partial coverage. Terminus-2 – Kimi K3 (always on), Claude Code – GLM-5.2 (max).

Data table: Output tokens per task

Output tokens per task
Model	Creator	Output tokens per task	Coverage (attempts)

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Terminus-2 – Kimi K3 (always on) or Claude Code – GLM-5.2 (max)?

Not comparable yet: Terminus-2 – Kimi K3 (always on) has partial coverage (GS-QA2 incomplete) and Claude Code – GLM-5.2 (max) has partial coverage (35 of 147 Earth-Bench V2 tasks; 165 of 173 GeoBenchX tasks; GS-QA2 incomplete), so it is not ranked. Terminus-2 – Kimi K3 (always on): 320 of 614 planned attempts. Claude Code – GLM-5.2 (max): 120 of 614 planned attempts (9 excluded). Coverage is partial; scores can change as more attempts are decided.

Which is cheaper per task, Terminus-2 – Kimi K3 (always on) or Claude Code – GLM-5.2 (max)?

Terminus-2 – Kimi K3 (always on), at $0.336 against $0.437 for Claude Code – GLM-5.2 (max).

Which writes fewer output tokens per task, Terminus-2 – Kimi K3 (always on) or Claude Code – GLM-5.2 (max)?

Terminus-2 – Kimi K3 (always on), at 9k against 34k for Claude Code – GLM-5.2 (max).

Which has the larger context window, Terminus-2 – Kimi K3 (always on) or Claude Code – GLM-5.2 (max)?

Terminus-2 – Kimi K3 (always on), at 1,000,000 tokens against 262,144 tokens for Claude Code – GLM-5.2 (max).

More: Terminus-2 – Kimi K3 (always on) · Claude Code – GLM-5.2 (max) · all comparisons