Connect to your AI assistant

One-time setup — then just ask Claude, Gemini, or ChatGPT to deploy your app.

Run this command once in your terminal

This registers Onairlo as an MCP server in Claude Code. You only need to do this once — it persists across all sessions.

              click to copy
              loading…
            

Start deploying

In any Claude Code conversation, mention @onairlo and describe what you want to deploy:

              @onairlo deploy this Flask app with MongoDB

              @onairlo deploy my React frontend

              @onairlo show logs for my app

Open Claude.ai Integrations settings

Go to claude.ai → Settings → Integrations and click Add Integration.

Enter the MCP server URL

Paste this URL into the integration URL field:

              click to copy
              loading…
            

Name it Onairlo and save. Claude.ai will ask you to authenticate — sign in with your Onairlo account.

Enable it in a conversation

Start a new chat, click the ⊕ Tools button, and toggle Onairlo on. Then ask it to deploy your app.

Copy your API key

Your Onairlo API key is in your . You'll need it in step 3.

Open Google AI Studio and add a tool

Go to aistudio.google.com → Settings → Tools → Add tool and select MCP Server. Paste this URL:

              click to copy
              loading…
            

Add your API key as a request header

In the MCP server config, add a custom header:

              Header name:  X-API-Key

              Header value: your-api-key

Save the tool — Gemini can now deploy and manage your apps.

Deploy from a Gemini conversation

Start a new chat in AI Studio or Gemini, enable the Onairlo tool, and ask it to deploy:

              Deploy this Flask app with MongoDB

              Show me my deployed apps

              Get logs for my app

Note: MCP tool support in Gemini requires Google AI Studio or the Gemini API with MCP integration enabled.

Copy your API key

Your Onairlo API key is in your . Copy it — you'll paste it into ChatGPT.

Add a custom GPT action

Go to ChatGPT → Explore GPTs → Create → Configure → Add Action. Set the schema URL to:

              click to copy
              loading…
            

Under Authentication, choose API Key, set header name to X-API-Key, and paste your key.

Deploy from your GPT

Save the GPT and ask it to deploy your app. It will call the Onairlo API automatically and return the live URL when ready.

Note: ChatGPT currently requires a ChatGPT Plus or Team subscription to use custom GPT actions.

Deploy in seconds

Ship apps with
one message to Claude

Tell your AI assistant to deploy your app. Onairlo provisions a real server, installs your stack, and returns a live HTTPS URL.

Claude Code

user → @onairlo deploy this Flask app with MongoDB

✓ Python runtime detected

✓ MongoDB dependency detected → provisioning on VM

✓ Dockerfile generated

◌ Building & deploying… (~3 min)

✓ Live at https://amber-river-12.onairlo.com

⚡

Zero config

Drop your files and go. Runtime, databases, and Dockerfile are all auto-detected from your project structure.

🗄️

Databases included

PostgreSQL, MySQL, MongoDB, and Redis are detected from your dependencies and provisioned on the VM automatically.

🔒

HTTPS out of the box

Every app gets a subdomain on onairlo.com with a wildcard TLS certificate. No DNS setup, no certbot commands.

🤖

Works with Claude, Gemini & ChatGPT

Connect once via MCP or the API — then deploy, update, check logs, and delete apps entirely from chat.

📦

Dedicated VMs

Every app runs on its own Hetzner VM — no containers shared with strangers, no cold starts, no resource contention.

📊

Built-in observability

Request stats, error logs, and container output available in the dashboard and via AI tools — no third-party setup.

How it works

From code to live URL in three steps

Connect once

Add Onairlo to Claude Code, Claude.ai, Gemini, or ChatGPT with a single command or URL using your API key.

Ask Claude to deploy

Say "@onairlo deploy this app" — Claude reads your files, detects the stack, and kicks off the deployment.

Get a live URL

A VM is provisioned, your app is built and started, and you get an HTTPS URL — no terminal, no cloud console.

Supported runtimes & databases

Node.js Python Go Ruby PHP Rust Java .NET Elixir Bun Deno Docker PostgreSQL MySQL MongoDB Redis

Ready to deploy?

Create a free account and connect to your AI assistant in under a minute.

Ship apps withone message to Claude

Sign up

Ship apps with
one message to Claude