Vaikora Documentation

The firewall for AI prompts. Vaikora sits between your app and the LLM and blocks prompt injection, data leaks, and jailbreaks in real time. Get protected in minutes.

Get a demo · Open-source gateway on GitHub · MCP server

Getting started

Introduction, quickstart, authentication, and your first protected request. Stand up enterprise-grade AI security in minutes with no change to your model provider.

Integration guides

Drop-in guides for Node.js, Python, React and Next.js, and Azure OpenAI, plus supported LLM providers and migration from the OpenAI SDK. Vaikora is an OpenAI-compatible proxy, so most apps integrate by changing one base URL.

Security and policies

Security overview, policy presets, custom security rules, and threat detection. Configure what blocks, redacts, or passes through for each environment.

Monitoring and analytics

Monitoring dashboard, usage tracking, spending and on-demand usage, and audit logs. See every request and every blocked threat.

Related pages