OpenClaw is the free, open-source AI agent that 250,000+ GitHub stars say is worth your time. This guide gets you from zero to running automated tasks in 30 minutes — even if you have never set up a developer tool before.
We will install Node.js, set up OpenClaw with a free API key, connect a Telegram bot, and run your first automated task. By the end, you will have a personal AI agent responding to your Telegram messages and ready for automation.
What You Need Before Starting
Before beginning, make sure you have:
- A computer running Windows 10+, macOS 12+, or Ubuntu 20.04+
- 30 minutes of uninterrupted time
- A stable internet connection (needed for initial downloads)
- A Telegram account (free — download the app if you do not have it)
- A free API key — we will use Groq's free tier (no credit card, just email signup)
Optional but recommended:
- 16GB RAM if you plan to run local Ollama models later
- A second terminal window or terminal split view
Not required:
- Programming experience
- Any existing AI subscriptions
- Payment information (we use free tiers in this guide)
Step 1: Install Node.js 22
OpenClaw requires Node.js 22 or later. Here is how to install it on each platform:
On macOS
# Install Homebrew if you don't have it (skip if already installed)
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
# Install Node.js 22 via Homebrew
brew install node@22
# Verify
node --version # Should show v22.x.x
npm --version # Should show 10.x.x or higher
On Ubuntu / Debian Linux
# Install nvm (Node Version Manager)
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.0/install.sh | bash
# Reload your terminal or run:
source ~/.bashrc
# Install Node.js 22
nvm install 22
nvm use 22
nvm alias default 22
# Verify
node --version
On Windows
- Visit nodejs.org and download the LTS installer (Node.js 22.x)
- Run the installer, accepting all defaults
- Open Command Prompt or PowerShell and verify:
node --version
npm --version
Step 2: Install OpenClaw
With Node.js installed, OpenClaw installs with a single command:
# Install OpenClaw globally
npm install -g openclaw
# Verify the installation
openclaw --version
You should see a version number like 0.x.x or 1.x.x. If you see an error, ensure Node.js 22+ is active with node --version and try again.
Step 3: Create Your Workspace
An OpenClaw workspace is the folder where your agent lives — its memory files, skills, cron configurations, and outputs are all stored here.
# Create a new workspace (replace "my-agent" with any name you like)
openclaw init my-agent
# Move into the workspace
cd my-agent
# See what was created
ls -la
You will see a folder structure like this:
my-agent/
├── config.yml # Main configuration file
├── ABOUT.md # Your profile (fill this in!)
├── MEMORY.md # Long-term memory (auto-populated)
├── memory/ # Daily memory logs
├── skills/ # Installed skills
├── crons/ # Scheduled task configurations
└── outputs/ # Where task results are saved
Immediately fill in your ABOUT.md file — this is the single most impactful 5 minutes in your OpenClaw setup:
# Open ABOUT.md in your text editor
# On macOS: open -e ABOUT.md
# On Linux: nano ABOUT.md
# On Windows: notepad ABOUT.md
Add your name, location (be sure to mention India), timezone (IST, UTC+5:30), work context, and preferences. This file shapes every response OpenClaw gives you.
Step 4: Configure Your API Key
OpenClaw needs an AI model to think. We recommend starting with Groq's free tier — it is fast, genuinely free (no credit card needed), and has generous limits for personal use.
Getting Your Free Groq API Key
- Go to console.groq.com
- Sign up with your email address (or Google/GitHub OAuth)
- Navigate to API Keys in the sidebar
- Click Create API Key
- Give it a name like "OpenClaw Personal"
- Copy the key (you will only see it once — save it somewhere)
Configuring OpenClaw With Your Groq Key
# Set Groq as your provider
openclaw config set model_provider groq
openclaw config set api_key gsk_YOUR_GROQ_KEY_HERE
openclaw config set model llama-3.1-70b-versatile
# Optional: Set a fast model for simple tasks
openclaw config set fast_model llama-3.1-8b-instant
Alternative: Use an Ollama Local Model (₹0, No API Key)
If you prefer running 100% locally with no external API:
# Install Ollama first
curl -fsSL https://ollama.com/install.sh | sh
# Pull a model (this downloads ~2GB)
ollama pull llama3.2:3b
# Configure OpenClaw to use Ollama
openclaw config set model_provider ollama
openclaw config set model llama3.2:3b
openclaw config set api_url http://localhost:11434
Step 5: Set Up Telegram Bot
Telegram is the recommended messaging channel for OpenClaw in India. Setting up a Telegram bot takes about 5 minutes.
Create Your Bot via BotFather
- Open Telegram and search for
@BotFather(the official bot for creating bots — blue checkmark) - Start the chat and send:
/newbot - BotFather will ask: "Alright, a new bot. How are we going to call it?" — Send a friendly name like
My OpenClaw Agent - Next it asks: "Now let's choose a username for your bot." — Send a unique username ending in
bot, likemyname_openclaw_bot - BotFather replies with your bot token — a string like:
7234567890:AAF_abc123def456ghi789jkl012mno345pqr - Copy this token — you need it now
Connect Telegram to OpenClaw
# Still in your workspace directory (my-agent/)
openclaw config set telegram_token YOUR_BOT_TOKEN_HERE
# Enable the Telegram channel
openclaw channel enable telegram
# Confirm the configuration
openclaw channel list
Step 6: Start OpenClaw and Test
You are ready to launch.
# Start OpenClaw (run this from your workspace directory)
openclaw start
You should see output like:
✓ OpenClaw v0.x.x starting...
✓ Workspace: /Users/yourname/my-agent
✓ Model: groq/llama-3.1-70b-versatile
✓ Memory: enabled
✓ Telegram: @myname_openclaw_bot connected
✓ Ready — send a message to begin
Send Your First Message
Open Telegram on your phone or desktop, search for your bot username (e.g., @myname_openclaw_bot), and send:
Hello! What can you do?
Within a few seconds, your OpenClaw agent will reply describing its capabilities. Congratulations — your AI agent is live.
Step 7: Send Your First Automated Task
Now let's try a real task that demonstrates what OpenClaw can actually do:
Task 1 — Summarise today's AI news:
Send this to your Telegram bot:
Search the web for the top 3 AI news stories from today and give me a 2-sentence summary of each.
Task 2 — Write a daily brief:
Create a brief daily agenda based on this: I have a 10 AM team standup, need to review 3 pull requests on GitHub, and want to finish a blog draft by 4 PM IST. Prioritise these and give me a structured plan for today.
Task 3 — Set up a recurring cron:
Send:
Create a morning cron that runs every weekday at 7:30 AM IST and tells me the weather in Bengaluru and a motivational quote for the day.
OpenClaw will create the cron configuration automatically. From tomorrow morning, you will get an automated 7:30 AM message without doing anything else.
Common Errors and Fixes
Error: openclaw: command not found
Node.js is installed but the global npm bin directory is not in your PATH.
# Add npm global bin to PATH
echo 'export PATH="$(npm prefix -g)/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc
openclaw --version
Error: API key invalid or 401 Unauthorized
Your Groq API key is incorrect or has been pasted with extra whitespace.
# View current config
openclaw config get api_key
# Reset it
openclaw config set api_key YOUR_EXACT_KEY_WITH_NO_SPACES
Error: Telegram bot not responding
The bot token may be incorrect, or the OpenClaw process stopped.
# Check if OpenClaw is running
openclaw status
# If not running, restart it
openclaw start
# Test the Telegram connection specifically
openclaw channel test telegram
Error: EACCES permission denied during npm install
On Linux/macOS, this means npm does not have permission to write to the global folder. Fix with nvm:
# Reinstall Node.js via nvm (nvm installs to user home, no sudo needed)
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.0/install.sh | bash
source ~/.bashrc
nvm install 22
nvm use 22
npm install -g openclaw
Ollama model is slow
If local Ollama responses are taking 30+ seconds, your model is too large for your RAM. Switch to a smaller model:
ollama pull llama3.2:3b # 2GB — fast on 8GB RAM
openclaw config set model llama3.2:3b
Next Steps
Your OpenClaw is running. Here is what to do next:
Immediately (today):
- Fill in your
ABOUT.mdwith your name, timezone, and work context - Install 2-3 skills for your most common tasks:
clawhub install daily-brief email-assistant-pro - Set up one morning cron to prove the automation value to yourself
This week:
- Read the OpenClaw Skills Complete Guide and install India-specific skills for IRCTC, GST, and NSE alerts
- Try the 21 OpenClaw Hacks to unlock Pearl memory and model routing
When you're ready:
- Set up local Ollama models for ₹0 ongoing cost (Step 4 alternative above)
- Move OpenClaw to a cheap VPS (₹800/month) for 24/7 availability even when your laptop is off
- Write your first custom skill for a workflow unique to your work
For full India-specific configuration including WhatsApp setup and Hindi prompts, see /learn/ai-personal-assistants/openclaw-setup-guide.