Metrion Documentation

The AI gateway that monitors, tracks, and optimizes your AI API usage across all providers.

Metrion is an AI gateway that sits between your application and AI providers like Anthropic, OpenAI, Gemini, Mistral, and Grok. Route your requests through Metrion to get real-time visibility into token usage, costs, latency, and errors — without changing your existing code.

Quick Start

Connect your first AI provider and send your first tracked request in minutes.

How It Works

Understand the proxy architecture and how Metrion tracks your usage.

Integration Guides

Step-by-step guides for Anthropic, OpenAI, Gemini, Mistral, and Grok.

API Reference

Full reference for all proxy endpoints, headers, and authentication.

Get started in 3 steps

1. Create your account

   Sign up at metrion.dev and get your Metrion API token (sk-metrion-xxx) from the Integration page.

2. Add your provider key

   Go to Integrate and add your API key for Anthropic, OpenAI, or any supported provider. Metrion stores it encrypted.

3. Point your app at Metrion

   Replace the base URL in your AI SDK with Metrion’s proxy endpoint. Your sk-metrion-xxx token is your new API key.

What you get

• Real-time dashboard — Monitor token usage, costs, latency, and error rates across all providers.
• Request logs — Browse every individual request with filtering, search, and CSV export.
• AI Insights — Get AI-generated recommendations to optimize your model usage and reduce costs.
• Alerts — Set thresholds on budget, error rate, latency p95, or request volume and get email notifications.
• Multi-provider — Anthropic, OpenAI, Gemini, Mistral, and Grok all in one place.