Announcing Built On Envoy: Making Envoy Extensions Accessible to Everyone

Learn more

An Enterprise AI Gateway For Your Agents, Models, & MCP Servers

Route to any model, attribute spend by team and agent, failover automatically, debug production issues in minutes, and enforce data protection policies. All without rewriting your agents.

AI Traffic Management for Models, Tools, and the Policies that Govern Them

AI Gateway

Routes all LLM traffic across your agents, handles provider failover, enforces token budgets, and produces unified logs across every team and provider. The core of Agent Router.

  • Multi-model routing with automatic failover
  • Per-team, per-agent token budgets enforced inline
  • Unified logs across all providers and frameworks
  • OpenAI-compatible API, works with existing code
  • Enterprise SSO for team and user attribution

MCP Gateway

Extends Agent Router to govern how agents connect to tools, services, and data sources. Admins curate what's available. Developers get access without custom integrations.

  • Curated MCP server catalogs set centrally, accessed instantly
  • Access profiles by team, role, or agent
  • Every tool call logged with the same attribution as LLM calls
  • Works with any MCP-compatible framework

AI Guardrails

Runtime policy enforcement across both model and tool traffic. PII redaction, prompt filtering, and transaction blocking before damage is done. Integrates with your existing guardrail providers.

  • PII detection and redaction on every request and response
  • Prompt and response filtering with custom rules
  • Kill-switch for misbehaving agents without touching code
  • Integrates with existing guardrail providers
  • Audit logs for every policy decision

Fast Developer Onboarding

  1. 01

    Point your agents at Agent Router

    Change one line. The base URL changes from the provider endpoint to Agent Router. Everything else stays the same,

  2. 02

    Configure routing and providers

    Add your model providers, set routing rules, configure failover targets. Takes minutes, not days.

  3. 03

    Every team gets the same on-ramp

    New developers get access through approved profiles. No ticket filing, no individual key setup, no inconsistent patterns.

  4. 04

    Add attribution, guardrails, and controls as you need them

    Start with routing. Add cost attribution, policy enforcement, and MCP governance when the org is ready. Enterprise capabilities activate without redeployment.

base_url=
"https://api.anthropic.com/v1"
base_url=
"https://your-org.agentrouter.tetrate.io/v1"

Change one line of code to get started

Runs Wherever Your Agents Run

Distributed gateway deployment for faster inference, region-specific models, provider-specific guardrails, or proximity to your data.

Managed Cloud

Tetrate-hosted. Fastest way to get started. No infrastructure to manage.

Enterprise

On-Premises

Deploy inside your own infrastructure. Data stays in your perimeter. Required for regulated industries.

Enterprise

Edge

Deploy edge inference by zip code or service area, with localized model catalogs and data controls.

Enterprise

Distributed across your footprint

Run gateways in your AWS, Azure, or Google Cloud VPC, managed by one control plane.

Works with What You Already Have

Agent Router is provider-neutral, framework-agnostic, and OpenAI-compatible. It doesn't replace your stack, it sits in front of it.

Model Providers

Self-hosted
models

Agent Frameworks

Any OpenAI-
compatible SDK

Observability

Any OTEL-
compatible
stack

Identity and Access

Any OIDC/SAML
provider
Any SSO provider

Start Fast. Scale When You're Ready.

Agent Router Service

FOR DEVELOPERS

Everything you need to get your agents routing through a gateway today, without a procurement conversation or IT ticket.

  • AI Gateway with multi-model routing and auto-failover
  • MCP Gateway — connect agents to tools securely
  • Your own token usage and cost logs
  • OpenAI-compatible API — works with your existing code

Agent Router Enterprise

FOR LEADERS MANAGING AI

Everything in Service, plus the visibility, attribution, and guardrails an engineering leader needs to run AI across multiple teams without losing track of what it costs or how it behaves.

  • Cross-team cost attribution, showback, and chargeback
  • Admin controls — model and MCP access profiles by team
  • Runtime AI Guardrails — PII redaction and policy enforcement
  • Enterprise SSO — every request carries authenticated identity
  • Distributed deployment — cloud, on-prem, edge, or per-region

Need More Help?

Work with Tetrate forward-deployed engineers to design safe agent operations

Talk To An Engineer