Skip to main content
Rhetoric Audit
v4.0.0
Test Results
About
Methodology ▾
Research
Log in
Add to Chrome
Benchmark Test Results
Compare FME prompt versions and model performance across all benchmark runs.
Test Versions
V20 — Unified Pipeline (93.3%)
New
V20 Guardrails Eval (100%)
3P
V19.1 — Ensemble (100%)
V19.0 — Bridge (85.7%)
V18.5
V18.4
V18.3