Terminus-2 – GLM-5.3 (low) vs. Codex – DeepSeek V4 Flash (high): geospatial agent comparison

Comparison between Terminus-2 – GLM-5.3 (low) and Codex – DeepSeek V4 Flash (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. Codex – DeepSeek V4 Flash (high)
Metric	Terminus-2 – GLM-5.3 (low)	Codex – DeepSeek V4 Flash (high)
Geospatial Agent Index	57*	46* (partial: attempt 1 on 112 of 565 tasks)
Earth-Bench V2 score	47	55
GeoBenchX score	52	37
GS-QA2 score	73	No data
GeoAgentBench score (guided GIS workflow; outside the index)	92	No data
GeoAnalystBench score (guided GIS workflow; outside the index)	95	No data
Raster and remote sensing score	65	54
Vector and overlay score	65	33
Networks and routing score	No data	No data
Climate and time series score	53	100
Spatial statistics and interpolation score	21	0
Mapping and cartography score	75	0
Recognising infeasible requests score	41	36
Cost per task	$0.285	$0.034
Time per task	6.5 min	8.9 min
Tokens per attempt	235k	398k
Turns per attempt	11.2	17.2
Attempts decided	565	65
Input price per 1M tokens	$1.40	$0.44
Output price per 1M tokens	$4.40	$1.32
Context window	1,048,576	1,048,576
Image input	No	No
Developer	Z.ai	DeepSeek

Charts

Partial coverage. Codex – DeepSeek V4 Flash (high): 11 of 147 Earth-Bench V2 tasks; 101 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

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	57*	No pending judgements; coverage incomplete; 180 results awaiting review; 5 trace reviews declined by the judge	3 of 3	565 of 859 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. Codex – DeepSeek V4 Flash (high).

Data table: Cost per task

Cost per task
Model	Creator	Cost per task	Coverage (attempts)
Terminus-2 – GLM-5.3 (low)	Z.ai	$0.285	565 of 859 planned attempts

Output tokens per task

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

Planned attempts still to come

How to read this chart

Not shown: partial coverage. Codex – DeepSeek V4 Flash (high).

Data table: Output tokens per task

Output tokens per task
Model	Creator	Output tokens per task	Coverage (attempts)
Terminus-2 – GLM-5.3 (low)	Z.ai	12k	565 of 859 planned attempts

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Terminus-2 – GLM-5.3 (low) or Codex – DeepSeek V4 Flash (high)?

Not comparable yet: Codex – DeepSeek V4 Flash (high) has partial coverage (11 of 147 Earth-Bench V2 tasks; 101 of 173 GeoBenchX tasks; GS-QA2 incomplete), so it is not ranked. Terminus-2 – GLM-5.3 (low): 565 of 859 planned attempts. Codex – DeepSeek V4 Flash (high): 65 of 614 planned attempts (97 excluded). Coverage is partial; scores can change as more attempts are decided.

Which is cheaper per task, Terminus-2 – GLM-5.3 (low) or Codex – DeepSeek V4 Flash (high)?

Codex – DeepSeek V4 Flash (high), at $0.034 against $0.285 for Terminus-2 – GLM-5.3 (low).

Which writes fewer output tokens per task, Terminus-2 – GLM-5.3 (low) or Codex – DeepSeek V4 Flash (high)?

Terminus-2 – GLM-5.3 (low), at 12k against 14k for Codex – DeepSeek V4 Flash (high).

Which has the larger context window, Terminus-2 – GLM-5.3 (low) or Codex – DeepSeek V4 Flash (high)?

They are level, at 1,048,576 tokens.

More: Terminus-2 – GLM-5.3 (low) · Codex – DeepSeek V4 Flash (high) · all comparisons