TL;DR
- What: Deploy a fully hosted OpenClaw AI bot — no server, no terminal required.
- How: Sign up on OneClaw → pick a template → paste your bot token → choose an AI model → click Deploy.
- Time: Under 60 seconds to a live bot.
- Cost: Plans start at $9.99/mo. See pricing.
- Platforms: Telegram and Discord supported out of the box.
What Is OneClaw and Why Use It for OpenClaw?
OpenClaw is a powerful, open-source conversational AI framework that lets you build and run AI-powered bots on virtually any messaging platform. It's flexible, extensible, and battle-tested — but self-hosting it traditionally means provisioning a server, configuring environment variables, setting up process managers, and keeping everything updated. For developers that's manageable; for everyone else, it's a wall.
OneClaw removes that wall entirely.
OneClaw is the official managed hosting platform for OpenClaw. It wraps the entire deployment and operations lifecycle into a clean dashboard experience. One-click deploy means exactly that: one click, and your OpenClaw instance is running in the cloud, monitored 24/7, and ready to serve conversations — in under 60 seconds.
If you want to go deeper into the infrastructure layer, check out the cloud deploy guide. But if you want to go fast, keep reading.
Prerequisites
Before you start, make sure you have the following ready. None of these require technical skills — they're just accounts and keys you'll gather from third-party services.
You will need:
- A OneClaw account — free to create at OneClaw.net. Plans start at $9.99/mo (see pricing).
- A bot token — from Telegram BotFather or the Discord Developer Portal, depending on which platform you want to deploy to.
- An AI model API key — from Anthropic (Claude), OpenAI (GPT-4o), Google (Gemini), or DeepSeek, depending on which model you want to power your bot.
- 5 minutes — seriously, that's all it takes.
That's the entire prerequisites list. No Docker, no cloud accounts, no CLI tools.
Step-by-Step: Deploying OpenClaw with One-Click
Step 1 — Sign Up for OneClaw
Head to OneClaw.net and create your account. The sign-up flow takes about 30 seconds. Once you're in, you'll land on the dashboard — your central hub for managing all OpenClaw deployments.
If you're evaluating plans before committing, the pricing page breaks down exactly what's included at each tier, including the number of concurrent bot instances, compute allocation, and support level.
Step 2 — Create a New Deployment
From the dashboard, click "New Deployment". You'll be prompted to give your deployment a name — this is just a label for your own reference, like `customer-support-bot` or `my-discord-assistant`.
Step 3 — Choose a Template
This is where OneClaw's template system shines. Instead of writing a configuration from scratch, browse the curated template library and pick one that fits your use case.
Available template categories include:
- Customer Support — handles FAQs, escalation flows, and ticket creation.
- Community Manager — moderation helpers, welcome messages, and member Q&A.
- Personal Assistant — scheduling, reminders, and general-purpose conversation.
- Developer Tools — code review prompts, documentation lookup, and CI/CD notifications.
- Custom — start from a blank slate and configure every parameter yourself.
Each template comes with a preview of the system prompt and default tool configuration. You can apply a template and then customize it further after deployment, all from the dashboard — no file editing required.
Tip: If you already have an OpenClaw config from a previous self-hosted setup, you can upload it directly and skip the template step entirely.
Step 4 — Configure Your Bot Token
Next, paste in your bot token. This is the credential that connects your OpenClaw instance to the messaging platform where your bot will live.
- Telegram: Get your token from @BotFather on Telegram. Send /newbot, follow the prompts, and copy the token it gives you. For the full walkthrough, see the Telegram setup guide.
- Discord: Go to the Discord Developer Portal, create a new application, add a Bot, and copy the token from the Bot settings page. Full instructions are in the Discord setup guide.
Paste the token into the Bot Token field in the OneClaw deployment form. OneClaw encrypts and stores it securely — you'll never need to paste it again unless you rotate the token.
Step 5 — Select Your AI Model
OneClaw supports four AI model providers out of the box:
| Model | Provider | Best For |
|---|---|---|
| Claude 3.5 Sonnet / Claude 3 Opus | Anthropic | Long-form reasoning, nuanced conversation |
| GPT-4o | OpenAI | General-purpose, broad capability |
| Gemini 1.5 Pro | Multimodal tasks, large context | |
| DeepSeek V2 | DeepSeek | Cost-efficient, strong coding tasks |
Select your preferred model from the dropdown and paste in the corresponding API key. If you want OneClaw to intelligently route conversations across multiple models — for example, using a cheaper model for simple queries and a more powerful one for complex ones — enable ClawRouters. ClawRouters is OneClaw's smart model routing layer, and it can significantly reduce your AI API costs without sacrificing response quality.
Step 6 — Deploy
Click "Deploy".
That's it. OneClaw provisions a containerized environment, injects your configuration, starts your OpenClaw instance, and runs an initial health check — all in under 60 seconds. When the status indicator turns green, your bot is live.
You'll see a deployment card appear on your dashboard with real-time status, uptime percentage, and a link to the instance logs.
Post-Deploy Configuration
Your bot is live, but there are a few things worth configuring in the first few minutes to make sure it behaves exactly how you want.
Connect Your Messaging Platform
If you haven't already, finalize the platform-side connection:
- Telegram: No extra steps — once your bot token is active, BotFather's webhook is automatically registered by OneClaw.
- Discord: You'll need to invite your bot to your server using the OAuth2 URL from the Discord Developer Portal, and grant it the appropriate permissions (Send Messages, Read Message History, Use Slash Commands). See the Discord setup guide for the exact permission integer and invite URL format.
Review and Customize Your System Prompt
Navigate to the Configuration tab of your deployment. Here you can edit the system prompt that shapes your bot's personality, tone, and behavior boundaries. Changes take effect on the next message your bot receives — no restart needed for prompt changes.
Set Environment Variables
If your template uses any external integrations (like a CRM webhook, a Notion database, or a custom API), add the relevant environment variables under the Environment tab. OneClaw stores these encrypted at rest.
Configure Alerts
Under Settings → Alerts, you can add an email address or webhook URL to be notified if your instance goes offline or if the auto-restart system is triggered. OneClaw performs health checks every 5 minutes, so in the worst case you'll know within 5 minutes if something is wrong — and it will have already been automatically restarted.
Troubleshooting Common Issues
Even with a one-click setup, occasionally things don't go as expected. Here are the most common issues and how to resolve them.
Bot token is invalid Double-check that you copied the full token with no trailing spaces. For Telegram, make sure you're using the token from BotFather, not an API ID or hash (those are different credentials). For Discord, make sure you're copying the Bot token, not the Client Secret.
Bot is online but not responding to messages
- Telegram: Confirm the bot hasn't been blocked. Try messaging it from a different account.
- Discord: Check that the bot has been invited to the correct server with the correct permissions. Also verify that "Message Content Intent" is enabled in the Discord Developer Portal for your application.
AI model is returning errors Check that your API key is valid and has sufficient credits/quota. Claude and GPT-4o keys can be verified in their respective provider dashboards. Quota errors will appear in your deployment's log stream, accessible from the dashboard.
Deployment is stuck in "Starting" state This occasionally happens if a health check fails during initialization. Click Restart on the deployment card. If the issue persists, check the logs for error output and consult the guides section or contact OneClaw support.
ClawRouter not activating ClawRouters require at least two model API keys to be configured. If you only have one key entered, the router will fall back to single-model mode automatically.
What Happens After You Deploy?
Once your OpenClaw instance is running on OneClaw, you get:
- Continuous uptime monitoring with health checks every 5 minutes and automatic restarts.
- Log streaming so you can watch conversations and debug issues in real time from the dashboard.
- Zero-downtime configuration updates — edit your system prompt or environment variables without restarting the bot.
- Usage metrics including message volume, response latency, and model token consumption.
- One-click scaling — need to handle more concurrent conversations? Upgrade your plan from the pricing page without any migration work.
All of this is managed for you. There's no infrastructure to babysit.
Next Steps
You've deployed OpenClaw. Here's where to go from here:
- Explore more templates at /templates to see what other configurations are possible.
- Read the full guides library at /guides for advanced topics like multi-bot setups, ClawRouter configuration, and custom tool integration.
- Set up your messaging platform with the Telegram setup guide or Discord setup guide.
- Go deeper on infrastructure with the cloud deploy guide if you want to understand what OneClaw is doing under the hood.
OpenClaw is a serious piece of software, and OneClaw makes it accessible to everyone — from first-time bot builders to teams running production AI workflows. The one-click deploy is just the beginning.