Skip to content

JailbreakZoo

A resource library on jailbreaks and defences for large models

Visit official site
Category AI models

What JailbreakZoo is

JailbreakZoo is described as a resource library on jailbreaks and defences for large language and vision-language models, aiming at a deeper understanding of AI security.

What you can do with it

  • Browse techniques by release timeline
  • Read matching defence methods
  • Add new research or documentation

Application scenarios

  • A researcher studies new weaknesses
  • A developer strengthens a product
  • A class learns model security

Getting started

  • Open the GitHub page
  • Read the introduction for context
  • Scan techniques and defences by timeline
  • Open a case for a specific model

Pros & cons

✓ What we like

  • Both attack and defence strategies are collected
  • Documentation improvements are welcomed from the community

! What to watch out for

  • The library is organised around published research
  • Use requires reading the GitHub material directly

FAQ

Which models are covered?

Large language models and vision-language models are named.

Are defences included?

Defence strategies for both model types are listed.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review