Skip to content

Your AI assistant, always on

OpenClaw is the open source assistant that actually does things: it clears your inbox, guards your calendar, books the travel, and reports back where you already chat. Paperclip.inc runs yours around the clock on EU infrastructure.

7 days free, no card. €10 per month after.

O OpenClaw Telegram online
Today
  • Good morning. Three things: your 10:00 moved to 11:30, I drafted replies to both invoice emails, and the Berlin train leaves at 17:12. Tickets are in your inbox.

    OpenClaw · 07:31
  • Book the usual hotel for Tuesday night.

    You · 08:02
  • Done. Hotel Oderberger, Tuesday night, €138. The confirmation is in your inbox and your calendar is blocked.

    OpenClaw · 08:03

Runs on the models you already use

  • Claude logo Claude
  • GPT logo GPT
  • Gemini logo Gemini
  • DeepSeek logo DeepSeek
  • Qwen logo Qwen

What it does

One assistant for the whole day.

You talk to it like a person. It quietly runs the small tasks that eat your week, and checks in when something needs you.

Getting started

Running in minutes.

01

Start your instance

Sign up and your dedicated OpenClaw boots in the EU. First week free, no card.

02

Connect your chat

Link Telegram, Slack, or Discord, then paste your model key. The key stays on your instance.

03

Say hello

Message it like a person. It remembers you, your preferences, and what you asked last week.

Why hosted

Open where it matters. Managed where it counts.

Open source, no lock-in

OpenClaw is open source. Audit it, fork it, or take it with you. We run the same code you can run yourself.

Always on, hosted for you

Your assistant runs around the clock on its own dedicated instance, with automatic updates and daily encrypted backups. No server to keep alive.

Private by design

One isolated instance per person on EU infrastructure, sandboxed with gVisor. Your conversations stay yours and never train anyone's models.

Your keys, your models

Paste your own model key during setup and pay your provider directly. The key stays on your instance, never in our database. Swap providers whenever you like.

Open source

The assistant is open source. We run it well.

Hosted OpenClaw is the same code you can read on GitHub, with the operations handled: updates, isolation, backups, and an instance that never sleeps. If you would rather run it yourself, the door is open.

Read the source on GitHub →

Pricing

One instance. One price.

€10 per month, or self-host free if you would rather run it yourself.

FAQ

Questions, answered.

What do I need to get started?

A chat app and a model key. Start your instance, link Telegram, Slack, or Discord, and paste a key from your model provider. Setup takes about five minutes.

Which models does it run on?

Claude, GPT, Gemini, DeepSeek, Qwen, and most other major providers. You bring your own key, pay your provider directly, and can swap models at any time.

Where does my data live?

On your own dedicated instance in the EU, on Hetzner infrastructure in Germany, isolated with gVisor and backed up daily with encryption. Your conversations never train anyone's models.

What happens after the free week?

Hosted OpenClaw is €10 per month. If you decide not to subscribe, your instance pauses and you can pick it back up later. No surprise charges.

Can I leave and take everything with me?

Yes. OpenClaw is open source, so you can export your data and run the same assistant on your own hardware. Cancel anytime.

Is this the same OpenClaw I can self-host?

Yes, the same open source code. We add the operations: a dedicated always-on instance, updates, backups, and EU hosting.

Get your assistant today.

Set up in minutes. First week free, no card.