helixordevelopers

Benchmarks

Measured results for Helixor components, with the hardware they ran on, every raw run, and the scripts to reproduce them. We publish absolute numbers only, and only numbers you can reproduce.

Headline results#

ComponentResultConditions
Decision Runtime, in-process24.9 µs p50, 48.8 µs p99 per evaluate() callRuntime 0.2.1 (Python build), built-in pack, Apple M5 Max, single process, load average 7.5 to 10.5
Decision Runtime, sidecar347 µs p50 per HTTP round trip on loopbackSame machine, one keep-alive connection

Coming next#

AreaWhat will be publishedStatus
Decision Runtime on server hardwareLinux x86_64 and aarch64, multi-process scaling, memory per engine, payload-size sensitivityPlanned
Native Decision RuntimeThe same suite against the compiled runtimePlanned
Rostering solverTime to a feasible roster on public rostering competition instances, with the dataset and scriptsPlanned
Hosted reasoningAnswer, abstain and refuse rates with calibration on a public, held-out question setPlanned

A result appears here only when its scripts, data and environment can be published with it. See Methodology.