How to Run MoltBot with DigitalOcean's One-Click Deploy

• Moltbot is a powerful, self-hosted AI assistant that can execute tasks like managing calendars, browsing the web, and running terminal commands. It should not be run on your personal machine due to significant security risks associated with giving an AI agent high-level system access.

• Deploying Moltbot on a cloud server provides a safer environment through security features like authenticated communication, hardened firewall rules, non-root user execution, Docker container isolation, and private Direct Message (DM) pairing.

• Moltbot is model-agnostic and supports various LLMs via Application Programming Interface (API) keys or local deployment, making it flexible for different use cases and preferences.

First, sign in to your DigitalOcean account and create a Droplet. On the Create Droplets page, under Region, select the region that best applies to you. Under Choose an Image, select the Marketplace tab.

In the search bar, type Moltbot and select the Moltbot image from the search results.

Next, choose a Droplet plan. The Basic plan with at least 4GB of RAM (such as the s-2vcpu-4gb size) is recommended for running Moltbot effectively.

Under Authentication, select SSH Key and add your SSH key if you haven’t already. If you need to create an SSH key, follow the instructions in How to Add SSH Keys to New or Existing Droplets.

Finally, give your Droplet a hostname (such as moltbot-server), and click Create Droplet.

Alternatively, you can create a Moltbot Droplet using the DigitalOcean API. To create a 4GB Moltbot Droplet in the NYC3 region, use the following curl command. You’ll need to either save your API access token to an environment variable or substitute it into the command:

curl -X POST -H 'Content-Type: application/json' \
     -H 'Authorization: Bearer '$TOKEN'' -d \
    '{"name":"choose_a_name","region":"nyc3","size":"s-2vcpu-4gb","image":"moltbot"}' \
    "https://api.digitalocean.com/v2/droplets"

Once your Droplet is created, it will take a few minutes to fully initialize. After initialization, you can SSH into your Droplet using the ipv4 address shown in your DigitalOcean dashboard:

ssh root@your_droplet_ip

Replace your_droplet_ip with your Droplet’s actual IP address.

Once logged in, the Moltbot installation will be ready to configure. The 1 Click Deploy Moltbot includes Moltbot version 2026.1.24-1 pre-installed with all necessary dependencies.

You will see a welcome message from Moltbot. Under the Control UI & Gateway Access section, you will see a Dashboard URL. Note the Dashboard URL value. You will use it later to access the GUI in your browser.

Within the terminal, choose Anthropic as your AI Provider. If you have access to Gradient AI, you can select that option. OpenAI models will be available soon. Once you select your provider, provide the respective API/Secret key.

This will set you up to start using Moltbot with your LLM.

There are two ways to use Moltbot: you can either use the Graphical User Interface (GUI) through your browser, or you can use the Text User Interface (TUI) through your terminal. To use the TUI, run the following command.

/opt/clawdbot-tui.sh

To use the GUI, open a browser and in the URL bar, paste the dashboard URL from earlier. This will open a standard Moltbot GUI directly in your browser that uses the Gateway token to authenticate you for additional security. This will direct you to the default Chat page.

Here you can type a message and send, and Moltbot will respond. For example, if you ask about what files it can see, it will tell you.

Input
What files can you currently see on my computer?

Output
Here’s a list of the files and directories currently visible in the sandbox workspace:
.
├── AGENTS.md
├── BOOTSTRAP.md
├── HEARTBEAT.md
├── USER.md
└── skills
    ├── 1password
    │   ├── SKILL.md
    │   └── references
...

From the GUI, you can review the bot’s usage, add communication channels, schedule cron jobs, add skills, and manage all aspects of Moltbot. You’ve now successfully deployed Moltbot on DigitalOcean and accessed it through a web browser. From here, you can explore additional Moltbot capabilities, such as browsing the web, managing files, or executing terminal commands on your Droplet.

Moltbot comes with over 50 skills automatically loaded in the skill registry. You can install skills in the GUI by navigating to the Skills section in the browser dashboard. For example, to integrate with Google Calendar, search for calendar, and click on `Install gog (brew).

A large number of skills are available to perform a wide range of tasks including managing your files, automating web browsing, monitoring health and smart home technologies, and managing social media communication. Read through What is Moltbot? for an overview of how Moltbot works and what Moltbot’s capabilities are.

Can I use a model other than Claude?

Yes, Moltbot is designed to be model-agnostic, so it does support models other than Anthropic’s Claude. It allows users to use various Large Language Models (LLMs) via API keys or locally. However, note that using the 1 Click Deploy Moltbot as outlined above, most users will only be able to use Anthropic (support for OpenAI coming soon!).

Can I deploy it on other operating systems that are not Linux?

Yes, you can deploy Moltbot on Windows, macOS, and Linux, and other platforms.

What are the main security concerns?

The main security concerns are its high-level system access, potential for misconfiguration, and its ability to execute arbitrary code that might be harmful to your system. It’s important to be aware of the environment in which it’s deployed and the access it has.

How do I give API Key access to my Agents?

It is possible to selectively give agents more control over the world around you. The default moltbot application will keep these keys together in an environment that is available to all agents. This configuration gives you control to inject the keys you want to the agents that should have those powers. On the “Agents” Menu bar, select the Agent you’d like to grant access (or “Defaults” for all), then under Sandbox > Docker > Env, add the select API Keys that should be used.

How does pricing work with Moltbot?

Moltbot is free and open-source to download and use, but you are paying for the LLM tokens. Therefore, the price depends on your usage. You should be careful with this because with scheduled jobs or other functionality, the costs can increase quickly and unexpectedly.

In this tutorial, you deployed DigitalOcean’s 1 Click Deploy Moltbot, creating a secure environment for your personal AI assistant. By running Moltbot on a cloud server instead of your local machine, you’ve significantly reduced security risks while maintaining the full functionality of this powerful tool.

The DigitalOcean 1 Click Deploy Moltbot provides critical security features out of the box—including authenticated communication, hardened firewall rules, Docker container isolation, and non-root user execution—that make it safer to experiment with AI agent capabilities. You accessed it through a web browser and can now execute various tasks through your preferred messaging apps. Next, try adding new skills to your Moltbot instance and customize the app to best suit your agentic needs.

• Moltbot Quickstart Guide
• How to Build Parallel Agentic Workflows with Python
• Mistral 3 Models on DigitalOcean