Terminus-2 – Llama 3.3 70B (Non-reasoning) vs. Codex – GPT-6 Luna (low): geospatial agent comparison

Comparison between Terminus-2 – Llama 3.3 70B (Non-reasoning) and Codex – GPT-6 Luna (low) 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 – Llama 3.3 70B (Non-reasoning) vs. Codex – GPT-6 Luna (low)
Metric	Terminus-2 – Llama 3.3 70B (Non-reasoning)	Codex – GPT-6 Luna (low)
Geospatial Agent Index	8*	51* (partial: attempt 1 on 449 of 488 tasks)
Earth-Bench V2 score	9	44
GeoBenchX score	6	50
GS-QA2 score	10	59
GeoAgentBench score (guided GIS workflow; outside the index)	8	92
GeoAnalystBench score (guided GIS workflow; outside the index)	0	79
Raster and remote sensing score	7	55
Vector and overlay score	12	57
Climate and time series score	10	49
Spatial statistics and interpolation score	0	38
Mapping and cartography score	2	71
Recognising infeasible requests score	13	30
Cost per task	$0.023	$0.0031
Time per task	3.1 min	1.4 min
Tokens per attempt	62k	114k
Turns per attempt	14.4	7.8
Attempts decided	488	449
Input price per 1M tokens	$0.29	$0.10
Output price per 1M tokens	$2.25	$0.50
Context window	24,000	No data
Image input	No	No data
Developer	Meta	OpenAI

Charts

Partial coverage. Codex – GPT-6 Luna (low): 111 of 150 GeoBenchX tasks. 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

Attempts still to run

Data table: Geospatial Agent Index

Geospatial Agent Index
Model	Creator	Geospatial Agent Index	Range while attempts are pending	Benchmarks covered	Coverage (attempts)
Terminus-2 – Llama 3.3 70B (Non-reasoning)	Meta	8*	No pending judgements; coverage incomplete; 23 results awaiting review	3 of 3	488 of 722 scheduled attempts

Cost per task

Average cost per task (USD) · Lower is better

Attempts still to run

How to read this chart

Not shown: partial coverage. 1 entries without a first attempt on every index task. What is still running is on the models page.

Data table: Cost per task

Cost per task
Model	Creator	Cost per task	Coverage (attempts)
Terminus-2 – Llama 3.3 70B (Non-reasoning)	Meta	$0.023	488 of 722 scheduled attempts

Output tokens per task

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

Attempts still to run

How to read this chart

Not shown: partial coverage. 1 entries without a first attempt on every index task. What is still running is on the models page.

Data table: Output tokens per task

Output tokens per task
Model	Creator	Output tokens per task	Coverage (attempts)
Terminus-2 – Llama 3.3 70B (Non-reasoning)	Meta	2k	488 of 722 scheduled attempts

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Terminus-2 – Llama 3.3 70B (Non-reasoning) or Codex – GPT-6 Luna (low)?

Not comparable yet: Codex – GPT-6 Luna (low) has partial coverage (111 of 150 GeoBenchX tasks), so it is not ranked. Terminus-2 – Llama 3.3 70B (Non-reasoning): 488 of 722 scheduled attempts. Codex – GPT-6 Luna (low): 449 of 722 scheduled attempts. Coverage is partial; scores can change as more attempts are decided.

Which is cheaper per task, Terminus-2 – Llama 3.3 70B (Non-reasoning) or Codex – GPT-6 Luna (low)?

Codex – GPT-6 Luna (low), at $0.0031 against $0.023 for Terminus-2 – Llama 3.3 70B (Non-reasoning).

Which writes fewer output tokens per task, Terminus-2 – Llama 3.3 70B (Non-reasoning) or Codex – GPT-6 Luna (low)?

Codex – GPT-6 Luna (low), at 1k against 2k for Terminus-2 – Llama 3.3 70B (Non-reasoning).

Which has the larger context window, Terminus-2 – Llama 3.3 70B (Non-reasoning) or Codex – GPT-6 Luna (low)?

Not known yet: Codex – GPT-6 Luna (low) has no context window data.

More: Terminus-2 – Llama 3.3 70B (Non-reasoning) · Codex – GPT-6 Luna (low) · all comparisons