AI-Native Cloud Platform.
Use your favorite AI coding agent to build, deploy and manage applications on Nodion.
Works with leading AI Coding Agents
Claude Code
Anthropic's agentic coding tool. Deploy to Nodion directly from your terminal with Claude.
OpenAI Codex
OpenAI's coding agent. Build and ship your projects to Nodion with Codex.
Gemini CLI
Google's AI coding agent for the terminal. Use Gemini CLI to deploy and manage apps on Nodion.
Three ways to connect.
Integrate your AI agent with Nodion through CLI, MCP or our REST API.
Install the Nodion CLI.
Download a pre-built binary and run nodion login.
Download Binaries
Grab a pre-built binary for your operating system, then run nodion login to authenticate.
Install via Terminal
Prefer the terminal? A single command detects your OS and architecture, fetches the matching release and drops the binary on your PATH.
macOS & Linux Bash
bashcurl -fsSL https://cli.nodion.com/install.sh | bash Installs to /usr/local/bin (falling back to ~/.local/bin) and runs chmod +x. Redirect with NODION_INSTALL_DIR=~/bin.
Windows PowerShell
powershellirm https://cli.nodion.com/install.ps1 | iex Installs to %LocalAppData%\Programs\nodion and adds that folder to your user PATH. No admin rights required. Override the target with $env:NODION_INSTALL_DIR='C:\Tools\nodion' before running.
Quickstart
bashnodion login
cd my-app
nodion link # pins this directory to a Nodion project + app
nodion deploy --wait # streams build/release logs until available nodion link writes a small nodion.yml into the current directory. From then on the CLI resolves the app automatically, so you don't need --app on deploy, env, logs or status. Perfect for AI agents working in a repo.
Deploying with a Git integration
For dockerfile and buildpacks apps, Nodion pulls from the branch you connected in the dashboard. The agent's workflow is: commit → push → nodion deploy --wait. The build runs on Nodion against the most recent commit of the connected branch, build logs stream into your terminal.
Connect to the hosted MCP endpoint
Nodion runs a Remote MCP server at https://mcp.nodion.com/mcp. Any MCP-compatible client can plug in, authenticate with a Bearer token, and drive your Nodion projects in natural language. No local daemon, no custom wiring.
1. Create an API key
Go to my.nodion.com → User Settings → Security, click Add API Key, pick the scopes you want to grant, optionally restrict the key to specific projects and pick an expiry. The token is shown only once, so copy it straight into your AI client's config.
2. Add Nodion to Claude Code
A single command registers the server. No config file to edit.
bashclaude mcp add --transport http nodion https://mcp.nodion.com/mcp \
--header "Authorization: Bearer YOUR_NODION_API_KEY" Start or restart a Claude Code session. The Nodion tools become available in chat right away.
Using a different client? The MCP docs have ready-to-paste configs for Zed, Codex, Gemini CLI, Cursor and Windsurf.
What your agent can do
Out of the box the agent gets read tools (apps.list, deployments.logs, env.list, …) and confirm-gated write tools (deploy, env.set, env.unset). Every tool call is audited on the Nodion side, so you can review what an agent did on your behalf.
Use the public REST API directly
Any AI agent that can make HTTP requests can drive Nodion. All endpoints live under https://api.nodion.com/v1/ and expect a bearer token. See the full API reference for every endpoint, payload and response.
1. Create an API key
Go to my.nodion.com → User Settings → Security, click Add API Key, pick the scopes you want and optionally restrict the key to specific projects. The token is shown only once, so copy it right away.
2. Make your first request
bashcurl https://api.nodion.com/v1/applications \
-H "Authorization: Bearer $NODION_TOKEN" 3. Trigger a deployment
bashcurl -X POST https://api.nodion.com/v1/applications/$APP_ID/deployments \
-H "Authorization: Bearer $NODION_TOKEN" \
-H "Content-Type: application/json" \
-d '{}' Full reference: API docs.
What your AI agent can do.
Full access to the Nodion platform, directly from your AI coding session.
Deploy
Push your code to production. Your AI agent can trigger deployments, monitor build progress and verify that everything is running.
Logs
Stream and analyze application logs in real-time. Your AI agent can read logs, identify errors and suggest fixes without leaving your coding session.
Metrics
Access CPU and RAM metrics for your applications. Your AI agent can detect performance issues and recommend scaling before your users notice.
Dockerfile
Nodion uses Dockerfiles to build your app. AI agents can analyze your project and generate an optimized Dockerfile tailored for Nodion.
Environment
Manage environment variables and secrets. Your AI agent can configure database connections, API keys and other settings for your app.
Scaling
Scale your application up or down. Your AI agent can adjust instance counts and instance types based on traffic patterns and metrics.
From code to production in seconds.
AI agents generate a Dockerfile for your project and deploy it to Nodion. That's it.
✓ detected Rails 7.1 · PostgreSQL 16 · Node 20 ✓ wrote Dockerfile ✓ wrote nodion.yml → my-api (project prod)
Deployment #42 created (status=waiting) ── status: building ── => FROM ruby:3.2-slim AS base => COPY Gemfile Gemfile.lock ./ => RUN bundle install --jobs 4 => COPY . . ── status: deploying ── rolling out to 2 instances… ✓ Deployment #42 finished: available live at https://my-api.nodion.app
Project: prod
Application: my-api available
strategy: dockerfile · 2 instances
Last deployment:
#42 available (user) "deploy"