Skip to content
Toolradar
Enoki logo

Enoki

Claimed

Red-team deployed AI agents, trace attacks, and verify security fixes

Free plan available
Freemium Security Vulnerability Scanning
Visit Website
Tracked since 2026

The Bottom Line

Best for
Engineering, AI and security teams shipping AI agents or MCP servers that need security, safety and policy testing on every release.
Price
Free plan available. Compare plans

Key facts

  • The first assessment of one agent is free, with no install, no SDK and no credit card.
  • Every finding shows the exact attack and the agent's response, mapped to OWASP LLM and Agentic Top 10 and MITRE ATLAS.
  • Enoki is available as a connector in the Claude directory, as an app in ChatGPT and in the official MCP Registry.

Not for

Teams that only want a runtime guardrail or prompt firewall in production, or that only benchmark a base model rather than a deployed agent.

What is Enoki?

Enoki is an AI red teaming platform that tests AI agents for security, safety and policy risks. It attacks the deployed agent through its endpoint the way an adversary would, over multiple turns, shows the exact attack and the agent's response behind every finding, and re-runs the attack to confirm a fix holds. Findings map to the OWASP Top 10 for LLM and Agentic Applications and MITRE ATLAS. It runs from a dashboard, a coding agent over MCP, or CI/CD, and also tests MCP servers.

Available on: Web, API

Data compiled by Toolradar How we evaluate
Source enokilabs.ai

Preview

Key Features

  • Adaptive security testing for AI agents
  • Safety testing for harmful content and advice
  • Reproducible findings with attack proof and severity
  • Maps findings to OWASP and MITRE ATLAS
  • Re-runs attacks on releases to verify fixes
  • REST API and MCP access for testing runs

Pricing Plans

Enoki plans and prices
Plan Price Details
Free

$0

The first assessment of one agent

  • No install, no SDK, no credit card
Standard

On request

For teams testing agents on every release

  • Slack and signed webhooks
Enterprise

On request

For organizations with audit and ticketing needs

  • Full evidence export, Jira, Linear and Git integration
View full pricing

Is Enoki worth the price?

Good value

Enoki's free tier is generous, offering a full assessment of one agent at no cost, which is ideal for initial evaluation.

The Standard and Enterprise tiers are priced on request, which suggests custom pricing that could be expensive for smaller teams but fair for organizations needing audit and ticketing integrations. This pricing is best for security teams that need to validate AI agent security without upfront commitment.

Full Enoki pricing: every plan, hidden costs and alternatives

Enoki by the numbers

Enoki compared with 737 Security tools on Toolradar, as of Nov 2026
Metric Enoki Peer median Position (1 is best)
Free plan Yes 49% have one n/a
Integrations listed 4 2 #5 of 92 (joint) tools with at least one integration recorded

Each position is counted among the tools that have data for that row, as noted under it. All Security data · How we compute this. Advertising and placement fees are not inputs to any position.

Cite this: Toolradar, “Enoki by the numbers”, Nov 2026.

Reviews

Improve Your Thinking Patterns Using ChatGPT cover
$99 Free with your review

Review Enoki, get a free AI guide

Share your experience and we will send you Improve Your Thinking Patterns Using ChatGPT, free.

Write a review

Best Enoki Alternatives

Top alternatives based on features, pricing, and user needs.

Lakera logo
Lakera Freemium

The AI-native security platform to accelerate GenAI development and protect against emerging threats.

HiddenLayer logo
HiddenLayer Paid

Secure your AI applications from discovery to runtime

Promptfoo logo
Promptfoo Freemium

Build secure AI applications with AI security testing integrated into your development workflow.

Mindgard logo
Mindgard Paid

Automated AI red teaming to find and fix security risks

Giskard logo
Giskard Freemium

Proactive AI red teaming and LLM security platform to prevent vulnerabilities in production.

See all security tools →

Still deciding?

Most buyers shortlist 2 or 3 tools before committing. Pull a side-by-side comparison or browse the full alternatives shortlist below.

Enoki vs Lakera Head-to-head: features, pricing, who wins Enoki vs HiddenLayer Head-to-head: features, pricing, who wins Enoki vs Promptfoo Head-to-head: features, pricing, who wins

Explore More

Enoki FAQ

How does Enoki test an AI agent that is already deployed?

Enoki attacks the deployed agent through its endpoint the way an adversary would, using multi-turn attacks rather than a single prompt. Every finding shows the exact attack and the agent's response, so reviewers can see precisely what happened. Enoki then re-runs the attack to confirm a fix holds.

How does Enoki's approach compare with Promptfoo?

Enoki is built around attacking a deployed agent through its endpoint over multiple turns, then re-running each attack to verify a fix holds. Findings are mapped to the OWASP Top 10 for LLM and Agentic Applications and MITRE ATLAS, and the same platform also tests MCP servers. Enoki can be driven from a dashboard, a coding agent over MCP, or CI/CD, so teams comparing it with Promptfoo should weigh those capabilities against their existing workflow.

What trade-offs should a buyer expect from Enoki?

Enoki works by attacking a deployed agent through its endpoint, so a reachable agent endpoint is a prerequisite for testing. Its output is a set of attack-backed findings that a team still has to triage and remediate, and the verification step only confirms a fix once the agent itself has been changed. Findings are organized around the OWASP Top 10 for LLM and Agentic Applications and MITRE ATLAS, which suits teams already working in those security frameworks.

Which teams get the most value from Enoki?

Enoki fits security teams responsible for AI agents and the AI engineering teams who build and ship those agents. It also suits teams running MCP servers, since Enoki tests MCP servers alongside deployed agents. Organizations that want security testing wired into CI/CD or driven from a coding agent over MCP are a natural fit as well.

How is Enoki priced?

Enoki is available on a free tier, with paid plans that add more usage and more features. Teams can start on the free tier and move to a paid plan as their testing volume and feature needs grow.

How does Enoki fit into a CI/CD pipeline or a coding-agent workflow?

Enoki runs from a dashboard, from a coding agent over MCP, or in CI/CD, so testing can happen wherever the team already works. The platform attacks the deployed agent through its endpoint over multiple turns and surfaces the exact attack and response behind each finding. Re-running an attack to confirm a fix holds can therefore be part of the same workflow that produced the change.

Can Enoki test MCP servers as well as AI agents?

Yes, Enoki tests MCP servers in addition to red-teaming deployed AI agents. Both are handled by the same platform, which attacks over multiple turns and shows the exact attack and the agent's response behind every finding. Results map to the OWASP Top 10 for LLM and Agentic Applications and MITRE ATLAS.

How does Enoki make each security finding actionable for reviewers?

Every Enoki finding shows the exact attack and the agent's response, so a reviewer can see what was sent and what came back. Findings map to the OWASP Top 10 for LLM and Agentic Applications and MITRE ATLAS, which gives teams a shared vocabulary for triage. After a fix is applied, Enoki re-runs the attack to confirm the issue is actually closed.

Source: enokilabs.ai

Guides & Articles

Best MCP Gateways & AI Agent Security Tools 2026

Expert guide

Best Deepfake Detection Tools 2026

Expert guide

Best AI Governance Tools 2026

Expert guide