How to Run OpenClaw for Under $20/Month

ByL.B. Kaelin

You want to run an AI agent that can code, browse, and automate tasks for you. The sticker shock from cloud bills is stopping you.

It doesn’t have to be expensive. You can run a fully functional OpenClaw agent for less than the cost of a streaming subscription. This guide is about getting real work done on a real budget.

By the end, you’ll have a specific, actionable plan to run OpenClaw for under $20 a month. You’ll know which services to use, where to cut costs without cutting corners, and how to keep it running reliably. Let’s get your agent online.

What You Need Before Starting

Think of this like packing for a camping trip. You need the right gear before you head into the woods. These are the non-negotiables. If you’re missing one, the rest won’t work.

  • A Computer or a Cloud Server: This is your campsite. You can use your own laptop (Linux, macOS, or Windows with WSL2) or a cheap cloud virtual machine. For under $20/month, a cloud VM is the most reliable option because it runs 24/7. We’ll pick a specific one later.
  • Docker and Docker Compose: This is your pre-fab tent and tools. OpenClaw runs in containers, which bundle all the software it needs. Docker is the system that runs these containers. Docker Compose is the instruction manual that tells them how to work together. Installing them is a one-time task.
  • A Code Editor (like VS Code): This is your headlamp. You’ll need to view and edit a few configuration files. A proper editor with syntax highlighting prevents simple typos that can waste hours.
  • An OpenAI API Key: This is your fuel. OpenClaw’s brain is a Large Language Model (LLM). You provide the key, and OpenAI charges you per use. The cost is minimal for light to moderate use, and you control the spending limit directly in your OpenAI account.
  • A GitHub Account: This is your map. You’ll clone the OpenClaw source code from GitHub. It’s also where you can report issues or contribute back later if you want.

Setting Up Your Environment

This is where you pitch the tent and unpack your gear. We’ll do it in a specific order so nothing depends on something that isn’t ready yet. If you’re using your own computer, follow the steps directly. If you’re using a cloud server, you’ll do this after you’ve created it.

1. Install Docker and Docker Compose

This is the foundation. Everything else runs on top of Docker. The commands differ by operating system. Pick your path.

For Debian/Ubuntu (or a cloud server):

# Update your package list
sudo apt update

# Install required packages for Docker's repository
sudo apt install -y ca-certificates curl gnupg

# Add Docker's official GPG key
sudo install -m 0755 -d /etc/apt/keyrings
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /etc/apt/keyrings/docker.gpg
sudo chmod a+r /etc/apt/keyrings/docker.gpg

# Set up the repository
echo \
 "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \
 $(. /etc/os-release && echo "$VERSION_CODENAME") stable" | \
 sudo tee /etc/apt/sources.list.d/docker.list > /dev/null

# Install Docker Engine, CLI, and Compose
sudo apt update
sudo apt install -y docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin

# Add your user to the 'docker' group so you don't need 'sudo' for every command
sudo usermod -aG docker $USER

# Apply the group change (you'll need to log out and back in for this to take full effect)
newgrp docker

After running these commands, verify the install with docker --version and docker compose version. You should see version numbers, not an error.

For macOS: Download and install Docker Desktop for Mac. It includes both Docker and Docker Compose. The installer handles everything.

For Windows: You must use Windows Subsystem for Linux 2 (WSL2). Install Docker Desktop for Windows, ensuring the “Use WSL 2 based engine” option is checked during installation. Then, open Ubuntu from your Start Menu and follow the Debian/Ubuntu commands above inside the WSL terminal.

2. Get Your OpenAI API Key

This step is done in your browser, not the terminal. Go to platform.openai.com/api-keys and create a new secret key. Copy it immediately. You won’t see it again.

Treat this key like a password. You will store it in a file on your machine, not hardcode it into scripts. We’ll do that next. While you’re in your OpenAI account, go to the Usage page and set a soft and hard spending limit. This is your ultimate cost control. Start with a $10 hard limit if you’re nervous. You can always raise it.

3. Clone the OpenClaw Repository

This downloads the latest OpenClaw code to your machine. It’s the blueprint for your agent.

# Clone the repository
git clone https://github.com/openclaw-ai/openclaw.git

# Move into the project directory
cd openclaw

You should now have an openclaw folder with files like docker-compose.yml and a .env.example file inside.

4. Configure Your Environment File

This is the most common point of failure. The .env file is where you store your secret API key and other settings. It’s ignored by Git, so your key stays local.

# Copy the example environment file to create your real one
cp .env.example .env

# Now, open the .env file in your code editor to edit it.
# For example, with VS Code:
code .env

Inside the .env file, find the line that says OPENAI_API_KEY=. Paste your secret key after the equals sign, with no spaces.

OPENAI_API_KEY=sk-your-actual-key-here

Save the file and close the editor. Your environment is now ready. The next steps involve choosing where to run it and starting it up.

Similar Posts