Benchmarks

Reproducible evaluation frameworks for generative search systems.

GSLHub benchmarks connect controlled prompts, repeated observations, transparent metrics and versioned datasets to make AI search visibility measurable and comparable.

GSL-BENCH-GEO-01VisibilityPilotv0.1.0

GSLHub Generative Search Visibility Benchmark

A reproducible benchmark designed to evaluate how generative search systems discover, include, cite and recommend digital sources across controlled prompts, repeated observations and structured visibility metrics.

Core metrics

AIRCRMCPRCR

Started Aug 14, 2026

Open specification

Specification, methodology and software are maintained independently.

Benchmark definitions are maintained in the public benchmarks repository, methodological protocols and coding rules in the research repository, and the reusable deterministic metric implementation in @gslhub/metrics-core.

Benchmark repository ↗Metrics software ↗Research methodology ↗