Quick Start

Pick an agent, launch it, connect your channels. About 60 seconds.

Prerequisites

That's it. No Node, no Docker, no SSH keys. The web dashboard handles everything.

Web Dashboard — the easy path

The fastest way to launch a bot is from the dashboard at launchpod.dev/account.

  1. Sign in — click "Sign Up" on the homepage and log in with email, Google, or GitHub.
  2. Start your trial — drop in a payment method. First month free, cancel anytime.
  3. Click "Create Instance" — pick a framework from the list of 4, confirm, and we'll spin up a dedicated VM in about 60 seconds.
  4. Open the browser terminal — once your instance shows "running", click the terminal button. An xterm.js session opens right in your browser — no SSH client needed.
  5. Run the onboard wizard — inside the terminal, follow the prompts to paste your LLM key and connect a messaging channel.

Supported Frameworks

LaunchPod supports 4 open-source agent frameworks. Pick one at instance creation — you can always spin up a new instance to try a different one.

FrameworkLanguageDescription
OpenClawNode.jsThe original. Personal AI with messaging, automation, browser, and cron.
NanobotPythonUltra-lightweight agent from HKUDS. 99% smaller than OpenClaw.
AstrBotPythonMulti-platform agentic chatbot with 1000+ plugins.
Hermes AgentPythonSelf-improving AI with persistent memory, from Nous Research.

OpenClaw ships with the latest 2026.4.8 release, which patches 8 critical CVEs and adds a bunch of new integrations.

CLI (optional, for power users)

Prefer the terminal? The LaunchPod CLI does everything the dashboard does, plus tailing logs and shell access. Install it with npm:

$ npm install -g @launchpod/cli

Then run:

$ launchpod init Opening browser for authentication... ✓ Signed in as you@email.com ? Pick a framework: ▶ OpenClaw Nanobot AstrBot Hermes Agent ◯ Provisioning... ✓ Instance ready. Connecting to terminal... Welcome to LaunchPod (Ubuntu 24.04) $

You now have a full bash shell on your VM. From here you can run the framework's onboard wizard and connect your channels.

Onboard Your Agent

Each framework has its own onboarding flow. For OpenClaw, the wizard runs automatically when you first connect:

✓ Welcome to OpenClaw! ⚠ Security: This agent can execute code. Accept? (y/n) ? Setup mode: QuickStart ? AI Provider: ▶ Anthropic (paste a setup-token from `claude setup-token`) OpenRouter (paste your API key) ? Messaging channel: ▶ Telegram (paste bot token from @BotFather) Discord ✓ Gateway service installed ✓ TUI launched — your bot is live!

For other frameworks, check that project's own README for setup — LaunchPod installs the framework cleanly on a fresh Ubuntu 24.04 VM, so any steps the upstream docs mention will work as-is.

Managing Your Instance

You can manage your instance from the web dashboard or the CLI — both hit the same API.

From the web dashboard

Head to launchpod.dev/account. You'll see your instance status, resource usage, and a terminal button that opens an xterm.js session right in your browser. No SSH setup, no local config.

From the CLI

If you closed your terminal, reconnect anytime:

$ launchpod connect <instance-name>

Check status and resource usage:

$ launchpod status <instance-name> Name: openclaw-1707654321 Status: running CPU: 12.3% Memory: 45.1%

Tail logs in real time:

$ launchpod logs <instance-name> -f

To disconnect from a terminal session, type exit or press Ctrl+D. Your instance keeps running in the background.

Plan & Specs

One plan, $9/mo, first month free. Every agent runs on a dedicated Hetzner Cloud VM.

vCPURAMStoragePriceIncludes
24 GB40 GB$9/mo1,000 free credits

Every instance runs Ubuntu 24.04 LTS with Node.js 22 + Python, Cloudflare tunnels (no exposed ports), automated daily backups, browser terminal, and CLI access.

Need more AI credits? Buy extra packs anytime — 500 for $5, 1,200 for $10, or 3,500 for $25.

Billing

Manage your subscription from the web dashboard or the CLI:

$ launchpod billing Billing portal opened in your browser.

From the Stripe billing portal you can:

If you cancel during the free first month, you pay nothing. After that, cancellations take effect at the end of your current billing cycle — your instance stays live until then.

Troubleshooting

Common issues and quick fixes.

Terminal connection keeps dropping

Symptoms: The browser or CLI terminal closes unexpectedly.

Fixes:

Can't connect to instance

Error: Terminal connection failed

Fixes:

CLI login fails or times out

Symptoms: launchpod login opens your browser but never completes.

Fixes:

Commands fail with "Unauthorized"

Error: Failed to authenticate. Please run: launchpod login

Fix: Your session probably expired (CLI tokens last 30 days). Re-authenticate:

$ launchpod logout $ launchpod login

Instance stuck in "provisioning"

Symptoms: Instance shows "provisioning" for more than 5 minutes.

Fixes:

"No active subscription" when creating an instance

Fix: Start your free trial from the account page:

$ launchpod billing

This opens Stripe checkout. After subscribing, try again.

Subscription is active but instance creation still fails

Fixes:

Terminal shows a blank screen

Fixes:

Still stuck?

Email hello@santuri.io — real humans respond. Include your instance name and we'll dig in.

Tip: keep the CLI updated

$ npm update -g @launchpod/cli

CLI Reference

All available commands in the LaunchPod CLI:

CommandDescription
launchpod initSign in, pick a framework, create an instance, and connect
launchpod connect <name>Open a terminal session on an existing instance
launchpod listList all your instances
launchpod status <name>Show instance details and resource usage
launchpod logs <name>View instance logs (-f to follow, -n for line count)
launchpod billingOpen the Stripe billing portal
launchpod loginAuthenticate with your LaunchPod account
launchpod logoutClear saved authentication token
launchpod whoamiShow the current authenticated user