Axis Spatial

Mistral Large 3 (Non-reasoning), Terminus-2 vs. GLM-4.7-Flash (Reasoning), Claude Code

Mistral Large 3 (Non-reasoning), Terminus-2 against GLM-4.7-Flash (Reasoning), Claude Code: Geospatial Agent Index, each evaluation, cost, time and tokens, on the same tasks and limits. The harnesses differ, so part of any gap can come from the harness. Methodology.

Side by side

Mistral Large 3 (Non-reasoning), Terminus-2 vs. GLM-4.7-Flash (Reasoning), Claude Code
MetricMistral Large 3 (Non-reasoning), Terminus-2GLM-4.7-Flash (Reasoning), Claude Code
Geospatial Agent Index: Text only33, not every task attempted yet41
Geospatial Agent Index: Text + visionStill to run: GS-QA2, Vision agentsNo image input
Earth-Bench V2 score3138
GeoBenchX score2939
GS-QA2 score3948
Vision agents scoreNo dataNo image input
GeoAgentBench score (Guided workflows; not in the Agent Index)46No data
GeoAnalystBench score (Guided workflows; not in the Agent Index)47No data
Raster and remote sensing score2941
Vector and overlay score4149
Climate and time series score3439
Spatial statistics and interpolation score823
Mapping and cartography score3349
Recognising infeasible requests score3340
Cost per task$0.026$0.27
Time per task (at 100 output tokens/s)1.1 min3.8 min
Tokens per attempt43k1.34M
Turns per attempt5.237.8
Attempts scored487488
Input price per 1M tokens$0.50$0.060
Output price per 1M tokens$1.50$0.40
Context window256,000131,072
Image inputYesNo
DeveloperMistral AIZ.ai

* Not every task attempted yet: the score may still change. What the asterisk means.

Charts

Not yet ranked. Mistral Large 3 (Non-reasoning), Terminus-2: 220 of 221 GS-QA2 tasks attempted. Grey placeholders mark these entries in the index, cost and output token charts; their height is not a value. Bars appear once every task has an attempt. Results so far are in the side-by-side table.

Geospatial Agent Index

Equal average of Earth-Bench V2, GeoBenchX and GS-QA2, 0 to 100 · Higher is better

Geospatial Agent IndexScores 0 to 100: GLM-4.7-Flash (Reasoning), Claude Code: 41; Mistral Large 3 (Non-reasoning), Terminus-2: Not yet ranked025507510041GLM-4.7-Flash (Reasoning)Not yetrankedMistral Large 3 (Non-reasoning)
Data table: Geospatial Agent Index
Geospatial Agent Index
ModelDeveloperGeospatial Agent IndexStatusEvaluations scoredTasks attempted
GLM-4.7-Flash (Reasoning), Claude CodeZ.ai41Complete3 of 3488 of 488 tasks (1 attempt not counted: provider or infrastructure error)

Cost per task

Average cost per task (USD) · Lower is better

Cost per taskCost per task in US dollars: GLM-4.7-Flash (Reasoning), Claude Code: $0.27; Mistral Large 3 (Non-reasoning), Terminus-2: Not yet ranked$0$0.10$0.20$0.30$0.40$0.27GLM-4.7-Flash (Reasoning)Not yetrankedMistral Large 3 (Non-reasoning)
How to read this chart

Not shown: not yet ranked. 1 entries without an attempt on every index task. What is still to run is on the models page.

Data table: Cost per task
Cost per task
ModelDeveloperCost per taskTasks attempted
GLM-4.7-Flash (Reasoning), Claude CodeZ.ai$0.27488 of 488 tasks (1 attempt not counted: provider or infrastructure error)

Output tokens per task

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

Output tokens per taskOutput tokens per task: GLM-4.7-Flash (Reasoning), Claude Code: 15k; Mistral Large 3 (Non-reasoning), Terminus-2: Not yet ranked05k10k15k20k15kGLM-4.7-Flash (Reasoning)Not yetrankedMistral Large 3 (Non-reasoning)
How to read this chart

Not shown: not yet ranked. 1 entries without an attempt on every index task. What is still to run is on the models page.

Data table: Output tokens per task
Output tokens per task
ModelDeveloperOutput tokens per taskTasks attempted
GLM-4.7-Flash (Reasoning), Claude CodeZ.ai15k488 of 488 tasks (1 attempt not counted: provider or infrastructure error)

Frequently asked questions

Which scores higher on the Geospatial Agent Index, Mistral Large 3 (Non-reasoning), Terminus-2 or GLM-4.7-Flash (Reasoning), Claude Code?
Not comparable yet: Mistral Large 3 (Non-reasoning), Terminus-2 is not yet ranked (220 of 221 GS-QA2 tasks attempted). Mistral Large 3 (Non-reasoning), Terminus-2: 487 of 488 tasks attempted. GLM-4.7-Flash (Reasoning), Claude Code: 488 of 488 tasks attempted. Scores can change as more tasks are attempted.
Which is cheaper per task, Mistral Large 3 (Non-reasoning), Terminus-2 or GLM-4.7-Flash (Reasoning), Claude Code?
Mistral Large 3 (Non-reasoning), Terminus-2, $0.026 to $0.27.
Which writes fewer output tokens per task, Mistral Large 3 (Non-reasoning), Terminus-2 or GLM-4.7-Flash (Reasoning), Claude Code?
Mistral Large 3 (Non-reasoning), Terminus-2, 4k to 15k.
Which has the larger context window, Mistral Large 3 (Non-reasoning), Terminus-2 or GLM-4.7-Flash (Reasoning), Claude Code?
Mistral Large 3 (Non-reasoning), Terminus-2, 256,000 tokens to 131,072 tokens.

More: Mistral Large 3 (Non-reasoning), Terminus-2 · GLM-4.7-Flash (Reasoning), Claude Code · All comparisons