Installing the Extension
Installing WebLLM takes less than 2 minutes. Follow these simple steps to get started.
System Requirements
Section titled “System Requirements”WebLLM works on any computer that runs Chrome:
- Browser: Chrome 120+ (or Edge, Brave, Arc - any Chromium browser)
- Operating System: Windows, macOS, or Linux
- For Local Models: 4GB+ RAM recommended
Installation Steps
Section titled “Installation Steps”Step 1: Open Chrome Web Store
Section titled “Step 1: Open Chrome Web Store”Click here to open WebLLM in Chrome Web Store
Or manually:
- Open Chrome
- Go to the Chrome Web Store
- Search for “WebLLM”
Step 2: Add to Chrome
Section titled “Step 2: Add to Chrome”- Click the “Add to Chrome” button
- A popup will appear showing the permissions WebLLM needs
- Click “Add Extension”
Understanding Permissions:
- Storage: Save your settings and API keys locally
- Web Requests: Communicate with AI providers you configure
- Scripting: Add the WebLLM API to websites
All data stays local. WebLLM has no backend servers. Learn more about privacy
Step 3: Pin the Extension
Section titled “Step 3: Pin the Extension”For easy access, pin WebLLM to your toolbar:
- Click the puzzle icon 🧩 in Chrome’s toolbar (top-right)
- Find WebLLM in the list
- Click the pin icon 📌 next to WebLLM
Now you’ll always see the WebLLM icon in your toolbar!
Step 4: Open WebLLM
Section titled “Step 4: Open WebLLM”Click the WebLLM icon in your toolbar. You’ll see the welcome screen.
✅ Installation complete!
First-Time Setup
Section titled “First-Time Setup”The first time you open WebLLM, you’ll go through a quick setup wizard.
Choose Your Provider Type
Section titled “Choose Your Provider Type”You’ll be asked to choose how you want to use AI:
Option 1: Free Local Model (Recommended)
Section titled “Option 1: Free Local Model (Recommended)”Best for: Getting started, privacy, offline use
- Click “Set up Local Model”
- You’ll see a list of available models:
- Llama 3.2 1B (1.2GB) - Fast, good for basic tasks
- Llama 3.2 3B (3.5GB) - Better quality, still fast
- Phi-3 Mini (2.3GB) - Great for code and technical content
- Click “Download” on your chosen model
- Wait for the download to complete (one-time only)
- Click “Activate Model”
Note: Local models download once and stay on your computer. You can add more models later.
Option 2: Use API Keys
Section titled “Option 2: Use API Keys”Best for: Premium AI capabilities (Claude, GPT-4)
- Click “Set up Cloud Provider”
- Choose your provider:
- Anthropic (Claude) - Get API key
- OpenAI (GPT-4) - Get API key
- Paste your API key
- (Optional) Set a spending limit
- Click “Save”
Don’t have an API key? That’s okay! You can:
- Use the free local model option instead
- Get an API key later (most providers offer free credits)
Option 3: Skip Setup
Section titled “Option 3: Skip Setup”You can skip setup and configure providers later from settings.
Test Your Setup
Section titled “Test Your Setup”After setup, try the built-in test:
- Click “Test AI” in the welcome screen
- Enter a simple prompt like “Say hello!”
- Click “Generate”
- You should see an AI response
✅ Setup complete!
Verify Installation
Section titled “Verify Installation”To make sure everything is working:
Check Extension is Active
Section titled “Check Extension is Active”- Look for the WebLLM icon in your toolbar
- It should be colorful (not gray)
- Click it - you should see the extension popup
Check Provider Status
Section titled “Check Provider Status”- Open WebLLM extension
- Go to “Providers” tab
- You should see at least one provider with a green checkmark
Test on a Website
Section titled “Test on a Website”Visit our demo page to test WebLLM with a real website:
The page will ask for permission to use WebLLM. Click “Allow” and try the AI features.
Updating the Extension
Section titled “Updating the Extension”WebLLM updates automatically through Chrome. To check for updates manually:
- Open
chrome://extensions - Enable “Developer mode” (top-right toggle)
- Click “Update” button that appears
- WebLLM will update if a new version is available
You’ll get a notification when updates are installed.
Troubleshooting Installation
Section titled “Troubleshooting Installation”Extension not showing up
Section titled “Extension not showing up”Solution:
- Check it installed: Go to
chrome://extensionsand look for WebLLM - Make sure it’s Enabled (toggle should be on)
- If not there, try installing again from Chrome Web Store
Can’t download local model
Section titled “Can’t download local model”Solutions:
- Check storage space: Local models need 1-4GB free space
- Check internet connection: Download requires stable connection
- Try a smaller model: Start with Llama 3.2 1B (1.2GB)
- Download later: You can use API keys first, download local models later
Extension icon is gray
Section titled “Extension icon is gray”A gray icon means WebLLM is installed but inactive.
Solutions:
- Click the icon to open the extension
- Make sure you have at least one provider configured
- Check
chrome://extensions- extension should be Enabled
API key not working
Section titled “API key not working”Solutions:
- Verify the key: Copy-paste carefully (no extra spaces)
- Check key permissions: Some API keys have restrictions
- Check account status: Make sure your provider account is active
- Try regenerating: Create a new API key from the provider
Permission popup not appearing
Section titled “Permission popup not appearing”Solutions:
- Check browser’s popup blocker settings
- Refresh the webpage
- Click the WebLLM icon manually to open it
Uninstalling WebLLM
Section titled “Uninstalling WebLLM”If you need to remove WebLLM:
Before Uninstalling
Section titled “Before Uninstalling”Consider:
- Export your data if you want to keep request history
- Save your settings (provider configs, API keys)
Uninstall Steps
Section titled “Uninstall Steps”- Go to
chrome://extensions - Find WebLLM
- Click “Remove”
- Confirm deletion
Note: This will delete all local data including:
- API keys
- Downloaded models
- Request history
- Settings
You can reinstall anytime from Chrome Web Store.
Next Steps
Section titled “Next Steps”Now that WebLLM is installed:
➡️ Configure Providers - Set up AI providers ➡️ Using WebLLM Websites - Learn how to use WebLLM on websites ➡️ Privacy & Data Control - Understand privacy features
Need Help?
Section titled “Need Help?”- FAQ - Common questions answered
- Troubleshooting - Fix common issues
- Community Forum - Get help from the community
- GitHub Issues - Report bugs
Ready to take control of AI? You’ve got WebLLM installed - now let’s configure it!