Enkrypt AI Guardrails

Enkrypt AI Guardrails is a runtime policy layer for AI agents, tools, retrieval systems, and model workflows. It is best suited to organizations securing an existing AI stack where controls and audit records need to span more than the final response. Its breadth is a compelling reason to consider it; credit allowances and retention on lower plans are meaningful constraints.

Overview

The product can approve, modify, or block risky behavior at prompt, retrieval, tool, and output boundaries. That scope matters for agent deployments: a risky action or data disclosure may happen in a tool call or retrieved material, not just in generated text. Coverage includes prompt injection, jailbreaks, unsafe tool actions, privilege or tenant-boundary violations, sensitive data exfiltration, toxic content, and compliance risks.

Policies also apply to image and audio inputs, including injection defenses. This makes Enkrypt relevant to multimodal workflows as well as text-based systems. The company was founded in 2022 and is headquartered in Brighton, Massachusetts.

Key features

Enforcement across the workflow

Filtering, rewriting, blocking, escalation, and tool-call approval or denial give teams several ways to respond to risk rather than relying on a single pass-or-fail filter. API-first integration, agent hooks or middleware, and an MCP Gateway offer routes into existing model and agent stacks. Identity claims can draw on Okta, Azure AD, or custom JWT claims, which is useful when policy decisions need identity context. Python is the listed SDK language.

Audit and latency

Enforcement logs include a policy ID, policy version, and reason code, and can be exported to a SIEM. That gives teams structured evidence for internal control review. The product page claims latency under 15 milliseconds and stable behavior under load; those claims make it worth evaluating for workflows sensitive to added delay, but do not remove the need to assess performance in the intended deployment.

The security posture lists SOC 2 Type II, ISO 27001, GDPR Ready, HIPAA Ready, and NIST AI RMF Aligned. These credentials and readiness statements may help teams screen vendors, while the policy controls and audit exports address a different need: governing runtime behavior.

Pricing

Explore is free, billed Free forever. It starts with 500 credits, then provides 50 credits per month, 7-day data retention, and community support. That is suitable for spot checks and initial evaluation, not a comprehensive red team assessment: that requires 5,000 credits.

Launch is $149 a month, billed monthly, with 5,000 credits to start, 250 credits per month, 30-day retention, and email support. Scale is $1,499 a month, billed monthly, with 10,000 credits to start, 1,000 credits per month, 30-day retention, and dedicated Slack or Teams support. Scale provides four times Launch's recurring monthly credits, but at a substantially higher monthly price; teams should compare that allowance with their expected usage.

The pricing note gives paid access from $134/mo (annual), while the Launch plan is shown at $149 a month billed monthly. Enterprise has custom pricing, custom allocation, custom retention, custom SLAs, and 24x7 support availability. It also includes VPC or on-premises deployment and unlimited usage for VPC, making it the clearest fit for organizations that need those deployment options or terms beyond the standard tiers.

Platforms

Enkrypt supports API, self-hosted, and web platforms. Enterprise adds VPC or on-premises deployment. The range suits teams choosing between integrating controls into a service and deploying within their own environment.

Who it's for

Enkrypt is aimed at enterprises securing AI deployments, AI safety researchers, and people building safer AI. It is strongest for teams managing agents, tools, retrieval, or multimodal inputs that need intervention points throughout a workflow and records of enforcement decisions. It is less suitable for a small-scale evaluation that needs more than 50 recurring monthly credits or longer than seven days of retention without moving to a paid plan.

Pros and cons

• Pros: Controls span prompts, retrieval, tools, and outputs, so teams can address risks beyond generated text.
• Pros: Multimodal policies, identity claims, and MCP support broaden its relevance to connected and varied AI workflows.
• Pros: Logs with policy IDs, versions, and reason codes can be exported to SIEM for internal control evidence.
• Cons: Explore's 50 monthly credits and 7-day retention are restrictive for sustained evaluation or review.
• Cons: The 5,000-credit threshold for a comprehensive red team assessment is far above Explore's recurring monthly allowance.
• Cons: Enterprise deployment options and custom terms require custom pricing, while the standard paid plans are billed monthly.

Alternatives

PromptGuard is a better fit when a reader prioritizes a free allowance of 20,000 scans per month, one API key, one project, and 24-hour log retention, or wants its listed extension and desktop platforms; its Team plan is $19 a month.

Amazon Bedrock Guardrails is an alternative for readers comparing a paid guardrails product with separate listed prices for text content filters and denied topics, and a free image content-filter plan.

Fiddler Guardrails is worth comparing for readers who want a free plan described as providing real-time detection of harmful exposure, hallucinations, toxicity, PII/PHI, prompt injection, and jailbreak attempts.

Openlayer Guardrails may suit readers who want a free starting tier with one member, five projects, one inference pipeline per project, and 20,000 inferences per month.

OpenSecureAI Prompt Firewall is an option for readers seeking a free plan with in-browser tools and npm packages, 2,000 API requests per month, and 30 requests per minute per key.

PII Firewall is a free, open-source choice for readers who specifically need a Linux or self-hosted Python package using regex-only installation without ML dependencies.

HiddenLayer AI Runtime Security is another paid runtime security alternative for readers comparing API, Linux, self-hosted, and web platforms.

Browse the AI Guardrail Software category to compare more products.

Verdict

Choose Enkrypt AI Guardrails if your organization needs runtime controls across agents, tools, retrieval, and multimodal inputs, with structured decisions it can export for audit evidence. Its main advantage is the breadth of enforcement points; its main drawback is that the free tier's recurring credits and retention are narrow, while deeper usage and private deployment lead to paid or custom-priced plans. For a small evaluation or a narrower, lower-cost need, compare alternatives before committing.