Skip to main content

0din.ai Docs

Welcome. Pick a documentation set to get started.

Defense NEW

Runtime protection: a fast, self-hostable classifier that scores every prompt for risk in real time, so you can block, log, or flag it before it reaches your LLM.

Paid version is self-hostable; hosted API in early-access beta.

AI Scanner

Open-source AI model security assessment platform. Run probes against your AI targets, score Attack Success Rate, and track trends over time.

Self-hostable. Apache 2.0 licensed.

0DIN Threat Feed

Browse the curated 0DIN catalog of disclosed AI vulnerabilities: reports, taxonomies, prompts & responses, industry variants, and detection signatures.

Free tier available; full content on Team and Enterprise.

More from 0DIN

  • 0din.ai: main site
  • Research: taxonomies, JEF, scoring systems, model cards