Evaluate AI · Documentation profile
Garak
LLM vulnerability scanner using probes and detectors to surface failure modes.
Primary documentation reviewed 2026-10-15. No hands-on testing, score or ranking.
Official project and documentation ↗When it fits
Run targeted checks against a model endpoint you control or are authorized to evaluate.
Limitations and setup
Scan coverage depends on the probes and detectors; a clean report is not a safety certification.
Suggested first trial
Select one relevant probe family and manually review flagged outputs.
Editorial suggestion; this experiment has not been run by Moyan.
An alternative to investigate
Inspect AIFramework for model evaluations composed from datasets, solvers and scorers.
Compare these workflows against your task; this is not a performance ranking.
Evidence and selection notes
Capabilities and setup notes are based on the linked primary documentation. Trial ideas and cautions about evaluation quality are editorial guidance. Check current licensing, model-provider charges and hardware requirements before adoption.
https://github.com/NVIDIA/garak