Frequently asked questions

ClawEasy is a managed hosting platform for OpenClaw, an open-source AI agent framework. We handle all the infrastructure — servers, SSL, DNS, software updates — so you can focus on using your AI agent, not setting it up.

You get a dedicated cloud server with OpenClaw pre-installed, connected to your messaging apps (Telegram, WhatsApp) and your AI provider of choice.

No. We designed ClawEasy specifically for non-technical users. If you can follow a setup guide and paste a few API keys into a form, you're all set.

Our step-by-step setup guide walks you through the entire process in under 15 minutes.

You need three things:

• A ClawEasy account (create one on this page)
• An API key from an AI provider — Anthropic, OpenAI, or Google (Google offers a free API tier; Anthropic and OpenAI are pay-as-you-go from the first request, but costs are very low for personal use)
• A Telegram account to create a bot (takes ~2 minutes, no coding required)

That's it. WhatsApp and Discord support is also available if you prefer those channels.

Most users are chatting with their AI agent within 10–15 minutes of signing up. Server provisioning takes about 3–5 minutes; the rest is pasting keys and connecting your messaging app.

Yes, for Starter, Standard, and Pro plans. Your ClawEasy subscription covers the hosting — the AI model usage is billed separately by the provider (Anthropic, OpenAI, or Google) based on how much you use it.

This keeps things transparent: you see exactly what you're using and paying for AI, with no markup from us.

Ultra plan: includes a local AI model running on the server itself — no external API key needed (though you can still use one if you prefer).

We currently support the following API providers and their latest models:

• Anthropic: Claude Sonnet 4.6, Claude Opus 4.6, Claude Haiku 4.5
• OpenAI: GPT-5, GPT-5 Mini, GPT-4o
• Google: Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 3 Flash
• Local (Ultra plan): DeepSeek R1 7B, Llama 3.1 8B, and other open-source models via Ollama — no API key needed

The model list is updated regularly. You can switch models at any time from your dashboard — no restart required.

This depends entirely on which model you use and how heavily. As a rough guide:

• Light use (a few dozen messages/day): usually <€2/month
• Moderate use: €3–10/month
• Heavy use with Claude Opus or GPT-4: can be more

Your dashboard shows live token usage and an estimated cost so you can track this.

On the Ultra plan, the local model runs with zero API cost — your AI usage is free, only electricity (covered by your subscription).

Yes. Your conversations go directly from your messaging app → your dedicated Instance → the AI provider you've chosen. ClawEasy does not intercept, read, log, or store your conversation content.

Your Instance is dedicated — not shared with other users. Your API keys are stored encrypted in our database.

All plans include a dedicated VPS, SSL, Cloudflare DNS, Telegram/WhatsApp support, and 20 TB monthly traffic. The differences are:

• Starter (CX22): 2 vCPU, 2 GB RAM — great for personal use with light to moderate conversation volume
• Standard (CPX22): 2 vCPU, 4 GB RAM — more headroom for heavier use or more complex tasks
• Pro (CPX32): 4 vCPU, 8 GB RAM — fastest response, best for power users
• Ultra (CPX41): 4 vCPU, 16 GB RAM — runs a full AI model locally on the server (no API costs)

Yes, you can upgrade or downgrade at any time from your dashboard. Upgrades take effect immediately (your server is migrated to a larger one). Billing is adjusted at your next renewal.

ClawEasy doesn't currently offer a free trial — a subscription is required to provision your server.

That said, you can keep costs low while evaluating: Google offers a free API tier through AI Studio (rate-limited but genuinely free), which is a good way to test the platform before committing to an AI spend. Anthropic and OpenAI are pay-as-you-go from the first request, with no free tier, but light personal use typically costs less than €2/month.

We accept all major credit and debit cards (Visa, Mastercard, Amex) through Stripe. Bank transfers and PayPal are not currently supported.

Your instance remains active until the end of the billing period. After that, it is shut down and deleted. We recommend exporting your data before cancelling — use the Export feature in your dashboard.

All instances run on Hetzner Cloud in Helsinki, Finland (EU). This means your data stays within the EU and is subject to GDPR-compliant infrastructure.

Not by default — your instance is managed by ClawEasy to ensure stability and security. Advanced users on Pro and Ultra plans can request SSH access by contacting us.

Yes. Your dashboard includes a file browser where you can upload, download, and delete files in a dedicated MyData folder on your instance. Your OpenClaw agent can read these files to help with tasks like document analysis, data lookup, and more.

When you upgrade or downgrade, your agent is migrated to a new server and the old one is deleted automatically. You are not charged for the rest of the month on the old server — Hetzner bills hourly, so you only pay for the exact hours the old server was running. There is a brief overlap of a few minutes during migration, which adds at most one extra hour at the old tier rate — typically just a few cents.

Your data is preserved during the migration. Your agent will be offline for approximately 10 minutes while the move happens.

This refers to the total data transferred in and out of your server. For an AI agent running on Telegram or WhatsApp, typical usage is a few gigabytes per month at most — you are extremely unlikely to come close to 20 TB. It's essentially unlimited for normal use.

If you do approach the limit (e.g., due to large file transfers), we'll notify you and your instance may be temporarily throttled.

OpenClaw supports custom tools and plugins, which you can configure through the agent setup. Direct shell access to run arbitrary scripts is not provided on standard plans.

Planned maintenance is rare and typically takes less than a minute (a service restart). We aim to do this outside peak hours. Hetzner infrastructure events are outside our control but equally rare. We do not offer an uptime SLA — see our Terms of Service for details.