Agentic infrastructure to power
AI-native companies.

Metorial gives every employee, agent, and AI client in your company one place to connect tools and skills. Governance, access control, and tracing included.

Connect any agent to any tool.

Flexible infrastructure for any agentic workflow.
Build customer-facing agents, internal assistants, and an AI workforce on a platform that scales with you. We build products to help teams become agent-native.

Built for devs.

Get started today.
Integrate tonight.

Metorial gives developers official TypeScript and Python SDKs to connect AI agents to tools in minutes.

TypeScript SDK

GitHub
npm install metorial @metorial/openai openai

Python SDK

GitHub
pip install metorial openai
Agent integration
1import Metorial from 'metorial';
2import { metorialOpenAI } from '@metorial/openai';
3import OpenAI from 'openai';
4
5let metorial = new Metorial({ apiKey: process.env.METORIAL_API_KEY! });
6let openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY! });
7
8let session = await metorial.connect({
9 adapter: metorialOpenAI.chatCompletions(),
10 providers: [
11 {
12 providerDeploymentId: 'github-deployment-id',
13 providerAuthConfigId: 'github-auth-config-id'
14 }
15 ]
16});
17
18await openai.chat.completions.create({
19 model: 'gpt-4o',
20 messages: [{ role: 'user', content: 'Summarize my open GitHub issues.' }],
21 tools: session.tools() // Pass tools to the agent.
22});
1import os
2from metorial import Metorial, metorial_openai
3from openai import AsyncOpenAI
4
5metorial = Metorial(api_key=os.environ["METORIAL_API_KEY"])
6openai = AsyncOpenAI(api_key=os.environ["OPENAI_API_KEY"])
7
8session = await metorial.connect(
9 adapter=metorial_openai(),
10 providers=[
11 {
12 "provider_deployment_id": "github-deployment-id",
13 "provider_auth_config_id": "github-auth-config-id",
14 }
15 ],
16)
17
18await openai.chat.completions.create(
19 model="gpt-4o",
20 messages=[{"role": "user", "content": "Summarize my open GitHub issues."}],
21 tools=session.tools(), # Pass tools to the agent.
22)

Connect any provider.

Connect anything.

Use 1000+ Metorial-provided integrations, or connect internal APIs, private tools, and remote MCP servers.

Ship instantly.

Launch in 3 clicks with OAuth out of the box. Cold starts under a second. True per-user isolation at scale.

Fully observable.

End-to-end tracing of every request, response, and error. Production-ready monitoring from day one.

Hundreds of integrations.
One simple API.

Give your agent superpowers with seamless connections to tools, data, and more-powered by MCP.

Monitor, optimize and debug.
Powerful tools for developers.

Skip the guesswork. Metorial provides you with the tools you need to monitor and analyze your agent's in every detail.

Monitor MCP Sessions

Track your MCP server's performance, find issues, and optimize your agents with ease.

Detailed Insights

Metorial provides the information you need to monitor how your agents interact with MCP servers hosted on Metorial Connect. Track usage, find issues, look at interactions between your agents and the MCP servers.

MCP Usage

Usage

5/13/2026

Detailed Logs

Metorial stores every MCP session, allowing you to replay and audit your agents' interactions with the MCP servers.

Received new message

Starting agent

Sending request to OpenAI

OpenAI Chat GPT

Using LLM to generate text.

OpenAI step completed successfully

OpenAI requested Weather information

Calling tool Weather

Weather

Retrieve weather information.

Weather step completed successfully

Sending weather information to OpenAI GPT

Monitor Issues

Find issues in your agents, and get notified when something goes wrong.

Received message from agent

Starting MCP server

Salesforce MCP

Invalid tool parameters

Reporting issue to agent

Stopping MCP server

Message Input

The actor takes a message as its input.

Code Action

Custom TypeScript code to transform the input.

AI Model

Prompt GPT-4o to generate a response.

Search Web

Search the web for additional information.

AI Model

Use a custom AI model to generate a response.

Text Response

Return the generated text back to your application.

Engineered for scale.
Built for enterprise.

Metorial is built for serious scale. Our proprietary hibernation technology ensures your integrations respond fast, without breaking the bank.

Limitless scaling

Scale from zero to hundreds of thousands of connections with millions of requests instantly.

Proprietary hibernation technology

We've built serverless for MCP. Our hibernation tech ensures your servers start in under a second, and stop when not in use.

Cheaper for everyone

Don't pay for connection duration. Only pay for requests served. Connect to Metorial for weeks, pay for seconds.

Built for enterprise

Whether you're a startup or a Fortune 500, Metorial scales with your business. Enterprise-grade scaling, support, and security come standard.

Bring any MCP server

Metorial makes every MCP server serverless. Our hibernation tech works with any server, custom or official. No code changes needed.

Security at scale

Metorial is built for security. Our infrastructure ensures true per-user isolation at scale, protecting your data and applications.

Go multi-tenant with ease

Metorial's architecture is designed for multi-tenancy. Effortlessly manage millions of users and connections with our robust platform.

Engineered for reliability

Sleep easy knowing Metorial's infrastructure is built for high availability and reliability. Your MCP servers stay online, no matter the scale.

Build your AI agent today. Add the tools and integrations you need. Get results right away. With Metorial.

Packed with Features

Metorial is a powerful AI integration platform that is designed to make it easy to build advanced agentic AI applications.

Fast response times

Secret vault

Error handling

Detailed logging and tracing

No MCP connection timeouts

First class SDKs

Ready to build with Metorial?

Connect AI to  

Built for developers

Get started in minutes

Serverless MCP servers

Usage based pricing

Open source

Based on the MCP standard

Star us on GitHub