What is Superagent?
Purpose-trained models for secure and compliant AI applications
Superagent provides three purpose-trained models — Guard, Verify, and Redact — available as standalone APIs that protect AI applications in real time. Each model is optimized for low-latency runtime protection in production environments.
Capabilities
Guard
Detects and blocks unsafe inputs, prompt injections, malicious tool calls, and backdoors before they reach your models
Verify
Validates model outputs against enterprise sources and policies to ensure every generation is factual and compliant
Redact
Removes sensitive data like PII, PHI, and secrets from text, logs, or documents automatically
Why Superagent
- Purpose-trained SLMs: Low-latency, optimized models for runtime AI protection
- Agent-native: Purpose-built to cover prompts, tool calls, function invocations, and file attachments
- Stops real threats: Detects and blocks prompt injections, malicious calls, data exfiltration, and backdoors
- Complete visibility: Stream structured logs, audits, and decisions into your existing security stack
Get Started
Ready to secure your AI applications? Get your API key and make your first call in minutes with our quickstart guide. Test Guard, Verify, or Redact via API, SDK, or CLI.
API
REST API for Guard, Verify, and Redact
SDK
Typed Python bindings with OpenAI-compatible APIs
TypeScript SDK
Drop-in client for Node.js, Vite, and edge runtimes
MCP
MCP server for your AI Agent
CLI
Command-line tool for testing and automation
Support
Need help or have questions? Connect with us:
- X (Twitter): Follow @superagent_ai for updates and announcements
- Discord: Join our community at discord.gg/spZ7MnqFT4 for support and discussions
- GitHub: Check out our source code and contribute at github.com/superagent-ai/superagent