Skip to main content
ZNYX AI
Open-source runtime · Hosted console for teams

Enterprise-grade AI guardrails for production LLM applications.

Run the open-source runtime in your own infrastructure. Add a hosted console when you need policy workflows, traces, and team operations.

Start FreeRead the Docs

Open-source runtime

Run guardrails in your own infrastructure with Docker, Kubernetes, or Helm.

Hosted console for Growth

Add policies, traces, and team operations without moving the runtime out of your environment.

Metadata-first visibility

Growth is designed around operational metadata by default rather than prompt and response bodies.

Optional full self-hosting

Enterprise can self-host both the runtime and the console with custom deployment support.

What is open source, and what is commercial?

The runtime and the control plane serve different roles. The runtime is the thing you deploy. The console is the thing your team uses to manage it.

Runtime

Open Source

The runtime is the data plane. It evaluates prompts, outputs, and streams in your infrastructure with no console dependency on Starter.

  • Docker, Kubernetes, and Helm deployment
  • Built-in detectors for PII, secrets, jailbreaks, toxicity, and more
  • Streaming evaluation, policy files, and SDK integration

Console

Growth / Enterprise

The console is the control plane. It helps teams manage policies, review traces, and operate guardrails across projects and environments.

  • Hosted in Growth, optionally self-hosted in Enterprise
  • Policy versioning, auditability, and operational views
  • Designed for metadata-first visibility by default

Buy the control plane only when the team needs it

ZNYX is not trying to move your runtime into a hosted black box. Growth is the step up for policy rollout, traces, and team coordination. Enterprise adds deployment flexibility for the console as well.

Protect prompts and outputs

Apply guardrails before the model call, after the model call, and during streaming responses.

Keep the runtime where your data lives

Every plan keeps the runtime in your infrastructure. Growth adds a hosted control plane instead of a hosted runtime.

Operate safely at team scale

Use policy versioning, environment-aware rollout, and operational traces when you outgrow local YAML-only workflows.

Operational view

Metadata-first by default

Stored by default in Growth
Trace id, policy decision, detector summary, latency, project scope, and operational metadata.
Not the default operating model
Full prompt and response body visibility in the hosted console.
Enterprise option
Self-host the console as well when deployment or procurement requires it.

How teams adopt ZNYX

Most teams do not need the same thing on day one. Start with the runtime, then graduate into the hosted control plane - or Enterprise self-hosting - when the team is ready.

Starter

Single team or product exploration using the open-source runtime only.

Growth

Self-hosted runtime with a hosted control plane for shared operations and governance.

Enterprise

Custom deployment models, including optional self-hosting for runtime and console.

Explore Deployment Patterns

Open-source runtime, hosted console

The runtime gives you self-hosted control from the beginning. The console adds operational workflows when you need shared policy management and cross-team visibility.

Included in the open-source runtime

  • Self-hosted runtime
  • Policy evaluation endpoints
  • Built-in detectors and streaming support
  • Python and TypeScript SDKs
  • Docker, Kubernetes, and Helm deployment

Available through the console plans

  • Hosted control plane in Growth
  • Policy management and rollout workflows
  • Operational traces and team workflows
  • Optional console self-hosting in Enterprise
  • Enterprise deployment and support options
Built-in detectors

Guardrails for the risks you actually ship against

Every detector runs in the self-hosted runtime. Compose them into policies, mix with custom detectors, and apply on input, output, or streaming responses.

PII

Input & Output

Detect and redact personally identifiable information across emails, phone numbers, addresses, IDs, and more.

Secrets

Input & Output

Catch leaked API keys, tokens, credentials, and other sensitive material before they reach a model or a downstream system.

Prompt injection

Input

Identify prompt-injection patterns and adversarial instruction chains aimed at overriding your application prompt.

Jailbreak

Input

Recognize jailbreak templates and evasion attempts that try to bypass policy controls.

Toxicity

Input & Output

Flag abusive, hateful, or harassing content across both user input and model output.

Profanity

Input & Output

Filter profanity with configurable severity thresholds per project and environment.

Topic control

Input & Output

Keep conversations within allowed topic boundaries or block specified off-topic areas.

Language

Input & Output

Enforce language restrictions and detect unexpected or unsupported languages in traffic.

Competitor mentions

Output

Prevent model responses from referencing competitor names or products you have excluded.

Hallucination signals

Output

Surface low-confidence or unsupported claims in responses for review or blocking.

Code safety

Output

Detect insecure code patterns, shell-escape risks, and dangerous suggestions in code responses.

URL and domain control

Input & Output

Block, allow, or inspect URLs and domains referenced inside prompts and responses.

View Detector Reference

Self-host the runtime in minutes — free, forever.

Pull the open-source runtime, drop it into your stack, and start enforcing policies. Add the hosted console when you want centralized policies, traces, and team workflows.

Start FreeReview Documentation