Claude Code – GLM-4.7-Flash (thinking on) vs. Codex – GPT-6 Sol (medium): geospatial agent comparison
Comparison between Claude Code – GLM-4.7-Flash (thinking on) and Codex – GPT-6 Sol (medium) on the Geospatial Agent Index, each benchmark, cost, time and token use, from the same tasks, agent and limits. Methodology.
Side by side
| Metric | Claude Code – GLM-4.7-Flash (thinking on) | Codex – GPT-6 Sol (medium) |
|---|---|---|
| Geospatial Agent Index | 41* | 77* (partial: attempt 1 on 193 of 488 tasks) |
| Earth-Bench V2 score | 38 | 80 |
| GeoBenchX score | 39 | 64 |
| GS-QA2 score | 48 | 88 |
| GeoAgentBench score (guided GIS workflow; outside the index) | No data | 100 |
| GeoAnalystBench score (guided GIS workflow; outside the index) | No data | No data |
| Raster and remote sensing score | 41 | 87 |
| Vector and overlay score | 49 | 86 |
| Climate and time series score | 39 | 83 |
| Spatial statistics and interpolation score | 23 | 50 |
| Mapping and cartography score | 49 | 100 |
| Recognising infeasible requests score | 40 | 50 |
| Cost per task | $0.27 | $0.048 |
| Time per task | 11.5 min | 1.3 min |
| Tokens per attempt | 1.34M | 87k |
| Turns per attempt | 37.8 | 6.5 |
| Attempts decided | 488 | 193 |
| Input price per 1M tokens | $0.060 | $2.00 |
| Output price per 1M tokens | $0.40 | $10.00 |
| Context window | 131,072 | No data |
| Image input | No | No data |
| Developer | Z.ai | OpenAI |
Charts
Partial coverage. Codex – GPT-6 Sol (medium): 55 of 117 Earth-Bench V2 tasks; 11 of 150 GeoBenchX tasks; 127 of 221 GS-QA2 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
First eight selected entries in metric order. Full results in table view.
Data table: Geospatial Agent Index
| Model | Creator | Geospatial Agent Index | Range while attempts are pending | Benchmarks covered | Coverage (attempts) |
|---|---|---|---|---|---|
| Claude Code – GLM-4.7-Flash (thinking on) | Z.ai | 41* | No pending judgements; coverage incomplete; 109 results awaiting review; 1 trace review declined by the judge | 3 of 3 | 488 of 722 scheduled attempts (1 excluded) |
Cost per task
Average cost per task (USD) · Lower is better
First eight selected entries in metric order. Full results in table view.
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
| Model | Creator | Cost per task | Coverage (attempts) |
|---|---|---|---|
| Claude Code – GLM-4.7-Flash (thinking on) | Z.ai | $0.27 | 488 of 722 scheduled attempts (1 excluded) |
Output tokens per task
Average output tokens per attempt, reasoning included · Lower is better
First eight selected entries in metric order. Full results in table view.
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
| Model | Creator | Output tokens per task | Coverage (attempts) |
|---|---|---|---|
| Claude Code – GLM-4.7-Flash (thinking on) | Z.ai | 15k | 488 of 722 scheduled attempts (1 excluded) |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Claude Code – GLM-4.7-Flash (thinking on) or Codex – GPT-6 Sol (medium)?
Which is cheaper per task, Claude Code – GLM-4.7-Flash (thinking on) or Codex – GPT-6 Sol (medium)?
Which writes fewer output tokens per task, Claude Code – GLM-4.7-Flash (thinking on) or Codex – GPT-6 Sol (medium)?
Which has the larger context window, Claude Code – GLM-4.7-Flash (thinking on) or Codex – GPT-6 Sol (medium)?
More: Claude Code – GLM-4.7-Flash (thinking on) · Codex – GPT-6 Sol (medium) · all comparisons