Experimental run record

Run 90a4a241

An inspectable execution record: configuration, deterministic outcomes, attempts, artefacts, and failure traces. Comparison applies across the configurations in this run only.

· pack geoworkbench-0.1 · accepted · sealed snapshot 3199b999

← All runs

Trials
12
Accepted
8
Configurations
4
Tasks
3
Cost USD
0.02152

In shortKimi K2.6 led this run: 2/3 tasks accepted (66.7%) for 0.0043 USD total. This ranking holds inside this run only — it is not a general model ranking.

Comparison across the configurations in this run only

JSON · CSV
Ranked by accepted rate, then total cost. Not comparable across runs or pack versions.
#ModelArchitectureAcceptedRateFirst-attempt successFalse-success rejectedCost USDCost / acceptedMedian msp95 ms
1Kimi K2.6single-shot2/366.7%66.7%10.00430.002131950 ms47663 ms
2GLM-4.7 Flashsingle-shot2/366.7%66.7%10.00440.002232054 ms55953 ms
3GLM-4.7 Flashsingle-retry2/366.7%66.7%10.00630.003226260 ms87049 ms
4Kimi K2.6single-retry2/366.7%66.7%10.00650.003244513 ms85304 ms

Trials

attempt traces expand beneath each trial
Trial outcomes are deterministic verifier results, not model-judge opinions.
TaskConfigurationRetryResultAttemptsCost USDLatency ms
crs-reproject-munichGLM-4.7 Flashsingle-shotaccepted10.0015 USD32054 ms
Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — accepted — cost 0.0015 USD — 32054 ms
# verifier: PASS — CRS reproject to WGS84 verified
# artifact: download (591 bytes)
import geopandas as gpd

# Load the input data
gdf = gpd.read_file("input.geojson")

# Diagnose CRS: Large coordinates indicate Web Mercator (EPSG:3857)
gdf.crs = "EPSG:3857"

# Reproject to WGS84 (EPSG:4326)
gdf_wgs84 = gdf.to_crs("EPSG:4326")

# Write the result
gdf_wgs84.to_file("output.geojson", driver="GeoJSON")
crs-reproject-munichGLM-4.7 Flashsingle-retryaccepted10.0013 USD26040 ms
Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — accepted — cost 0.0013 USD — 26040 ms
# verifier: PASS — CRS reproject to WGS84 verified
# artifact: download (591 bytes)
import geopandas as gpd

# Load the input file
gdf = gpd.read_file("input.geojson")

# Diagnose and assign CRS
# The prompt indicates coordinates are large (millions), indicating a projected CRS.
# We are instructed to determine the CRS is EPSG:3857.
gdf.crs = "EPSG:3857"

# Reproject to WGS84 (EPSG:4326)
gdf_wgs84 = gdf.to_crs("EPSG:4326")

# Write the result to output.geojson
gdf_wgs84.to_file("output.geojson", driver="GeoJSON")
raster-zonal-meanGLM-4.7 Flashsingle-shotrejected10.0019 USD55953 ms
Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — rejected — cost 0.0019 USD — 55953 ms
# verifier: FAIL — stats.json unreadable
# no artifact produced

raster-zonal-meanGLM-4.7 Flashsingle-retryrejected20.0038 USD87049 ms
Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — rejected — cost 0.0019 USD — 38014 ms
# verifier: FAIL — stats.json unreadable
# no artifact produced

Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — rejected — cost 0.0019 USD — 87049 ms
# verifier: FAIL — stats.json unreadable
# no artifact produced

vector-buffer-1kmGLM-4.7 Flashsingle-shotaccepted10.0010 USD23259 ms
Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — accepted — cost 0.0010 USD — 23259 ms
# verifier: PASS — 1000m buffers verified
# artifact: download (5935 bytes)
import geopandas as gpd

gdf = gpd.read_file("sites.geojson")
gdf_projected = gdf.to_crs(epsg=3857)
gdf_projected['geometry'] = gdf_projected.buffer(1000)
gdf_final = gdf_projected.to_crs(epsg=4326)
gdf_final.to_file("buffers.geojson", driver="GeoJSON")
vector-buffer-1kmGLM-4.7 Flashsingle-retryaccepted10.0012 USD26260 ms
Attempt trace undefined
# attempt undefined — model: @cf/zai-org/glm-4.7-flash — exit 0 — accepted — cost 0.0012 USD — 26260 ms
# verifier: PASS — 1000m buffers verified
# artifact: download (5903 bytes)
import geopandas as gpd

# Read the input file
gdf = gpd.read_file("sites.geojson")

# Reproject to EPSG:3857 (Web Mercator) for accurate 1000m buffering
gdf_projected = gdf.to_crs(epsg=3857)

# Create 1000m buffers
gdf_projected['buffer'] = gdf_projected.geometry.buffer(1000)

# Extract the buffer geometries and reproject back to EPSG:4326
buffers_gdf = gpd.GeoDataFrame(geometry=gdf_projected['buffer'], crs="EPSG:3857")
buffers_gdf = buffers_gdf.to_crs(epsg=4326)

# Write the output
buffers_gdf.to_file("buffers.geojson", driver="GeoJSON")
crs-reproject-munichKimi K2.6single-shotaccepted10.0012 USD31950 ms
Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — accepted — cost 0.0012 USD — 31950 ms
# verifier: PASS — CRS reproject to WGS84 verified
# artifact: download (591 bytes)
import geopandas as gpd

gdf = gpd.read_file("input.geojson")
gdf = gdf.set_crs(epsg=3857, allow_override=True)
gdf = gdf.to_crs(epsg=4326)
gdf.to_file("output.geojson", driver="GeoJSON")
crs-reproject-munichKimi K2.6single-retryaccepted10.0017 USD44513 ms
Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — accepted — cost 0.0017 USD — 44513 ms
# verifier: PASS — CRS reproject to WGS84 verified
# artifact: download (591 bytes)
import geopandas as gpd

gdf = gpd.read_file("input.geojson")
gdf = gdf.set_crs(epsg=3857, allow_override=True)
gdf = gdf.to_crs(epsg=4326)
gdf.to_file("output.geojson", driver="GeoJSON")
raster-zonal-meanKimi K2.6single-shotrejected10.0019 USD47663 ms
Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — rejected — cost 0.0019 USD — 47663 ms
# verifier: FAIL — stats.json unreadable
# no artifact produced

raster-zonal-meanKimi K2.6single-retryrejected20.0038 USD85304 ms
Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — rejected — cost 0.0019 USD — 38181 ms
# verifier: FAIL — stats.json unreadable
# no artifact produced

Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — rejected — cost 0.0019 USD — 85304 ms
# verifier: FAIL — stats.json unreadable
# no artifact produced

vector-buffer-1kmKimi K2.6single-shotaccepted10.0011 USD30670 ms
Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — accepted — cost 0.0011 USD — 30670 ms
# verifier: PASS — 1000m buffers verified
# artifact: download (5935 bytes)
import geopandas as gpd

gdf = gpd.read_file("sites.geojson")
gdf = gdf.set_crs("EPSG:4326", allow_override=True)
gdf_proj = gdf.to_crs("EPSG:3857")
gdf_proj["geometry"] = gdf_proj.buffer(1000)
gdf_out = gdf_proj.to_crs("EPSG:4326")
gdf_out.to_file("buffers.geojson", driver="GeoJSON")
vector-buffer-1kmKimi K2.6single-retryaccepted10.0010 USD24427 ms
Attempt trace undefined
# attempt undefined — model: @cf/moonshotai/kimi-k2.6 — exit 0 — accepted — cost 0.0010 USD — 24427 ms
# verifier: PASS — 1000m buffers verified
# artifact: download (5935 bytes)
import geopandas as gpd

gdf = gpd.read_file("sites.geojson")
if gdf.crs is None:
    gdf.set_crs(epsg=4326, inplace=True)

gdf_proj = gdf.to_crs(epsg=3857)
gdf_proj["geometry"] = gdf_proj.buffer(1000)

gdf_wgs84 = gdf_proj.to_crs(epsg=4326)
gdf_wgs84.to_file("buffers.geojson", driver="GeoJSON")

Cite this record

Axis Spatial. "Experimental run 90a4a241." Axis Spatial Arena, run 90a4a241-8f55-4b8c-921c-9ddfaa8f2ea2. Benchmark pack geoworkbench-0.1. https://www.axisspatial.com/arena/runs/90a4a241-8f55-4b8c-921c-9ddfaa8f2ea2 (accessed 19 Sep 2026).
@misc{axis-spatial-arena-run-90a4a241,
  author = {{Axis Spatial}},
  title = {Experimental run 90a4a241},
  howpublished = {Axis Spatial Arena, run: https://www.axisspatial.com/arena/runs/90a4a241-8f55-4b8c-921c-9ddfaa8f2ea2},
  note = {Benchmark pack geoworkbench-0.1},
  year = {2026}
}

Glossary

full protocol in the methodology
Pack version
The pinned set of task definitions a run executed. Results from different pack versions are never compared directly.
Configuration
A complete comparison unit: model, prompt version, tools, retry policy, and pinned task, data, and verifier versions. A model name alone is not comparable.
Lane
One configured route through which a model is reached. Duplicate upstream models across lanes count as adapter evidence, not extra capability.
False-success rejection
A trial whose output looked plausible but failed independent deterministic checks, such as an empty file or a fabricated value.
Token-cost proxy
Cost estimated from reported token usage and list prices at recording time; actual billing may differ.
Sealed snapshot
An immutable, content-addressed copy of a leaderboard computed from a finished run.