BMS

Start typing to search the publication.

Shareable benchmark table

GLM-5.3 arm wall time, tools, and localisation

Operational aggregates across all terminal runs of the GLM-5.3 arm. These are operational observations only and do not measure finding quality.

Table 18 · Operational telemetry

GLM-5.3 arm wall time, tools, and localisation

Agent system · GLM-5.3 + ZCode CLIn = 1 run / cellOperator contamination disclosedDeterministic evaluatorNot a model leaderboard
Operational aggregates across all terminal runs of the GLM-5.3 arm. These are operational observations only and do not measure finding quality.
MetricValue
Median wall time / run0m 09s
Total collection wall time5m 11s
Generated tokens (total)n/a (no token metering)
Tool calls (total)15
Source localised (tool-enabled runs)6/6
Table 18Operational aggregates across all terminal runs of the GLM-5.3 arm. These are operational observations only and do not measure finding quality.

Wall time includes GLM-5.3 + ZCode inference between turns and is not comparable to the parent pilot's automated local loop. Token telemetry does not exist for this runtime and is shown as n/a, never zero.

BMSbms.kyokasuigetsu.xyz
Download CSVDownload PNGShare table
Static image: https://bms.kyokasuigetsu.xyz/tables/glm-operational.png

Read the source publication