🧠 Ollama Setup Guide

Complete guide to set up Ollama backend for the Chrome Extension

✅

1. Install Ollama

Visit the official website to download Ollama:

👉 Download Ollama

Once installed, start the Ollama server:

ollama serve

This launches the Ollama API at:

http://localhost:11434
💡 Keep this running while using the extension.
🤖

2. Pull a Model (e.g. Gemma 3B)

After installation, run a model of your choice:

ollama run gemma3:1b

Once downloaded, you're ready to chat!

💡 Replace gemma3:1b with other models like llama3, mistral, etc.
🔐

Do You Need to Set Up CORS?

Yes, if you're using Firefox. Probably not for Chrome or other Chromium browsers (click to expand)

Whether you need to configure CORS depends on your browser:

✅ Chrome / Chromium (e.g., Brave, Edge)

If you're using Chrome-based browsers and extension version 0.1.3 or later, you likely do not need to set any CORS headers.

Ollama Client uses Chrome's Declarative Net Request (DNR) API to rewrite Origin headers in requests to localhost, which lets it bypass CORS errors without backend changes.

🦊 Firefox

Firefox does not support Chrome's DNR API, so manual configuration is required.

If you're using Firefox, set this in your environment:

OLLAMA_ORIGINS=moz-extension://*

When should you manually set OLLAMA_ORIGINS?

  • You're using Firefox
  • You're on extension version < 0.1.3
  • You're calling from localhost:3000 or another frontend
  • You're still getting ❌ 403 Forbidden errors

To cover both Chrome and Firefox, you can combine origins:

OLLAMA_ORIGINS=chrome-extension://*,moz-extension://*
🚫

3. Fix ❌ 403 Forbidden: CORS Error

If you're seeing CORS errors or using Firefox, follow these platform-specific instructions to set OLLAMA_ORIGINS:

macOS (Launch Agent)

1. Edit the plist file:

nano ~/Library/LaunchAgents/com.ollama.server.plist

2. Add inside <key>EnvironmentVariables</key>:

<key>OLLAMA_ORIGINS</key> <string>chrome-extension://*,moz-extension://*</string>

3. Reload the agent:

launchctl unload ~/Library/LaunchAgents/com.ollama.server.plist launchctl load -w ~/Library/LaunchAgents/com.ollama.server.plist

Linux (systemd)

1. Edit the service file:

sudo systemctl edit --full ollama.service

2. Add under [Service]:

Environment="OLLAMA_ORIGINS=chrome-extension://*,moz-extension://*"

3. Reload and restart:

sudo systemctl daemon-reload sudo systemctl restart ollama

Windows

1. Press Win + R, type sysdm.cpl, press Enter.

2. Go to Advanced → Environment Variables.

3. Add a new User variable:

  • Name: OLLAMA_ORIGINS
  • Value: chrome-extension://*,moz-extension://*

4. Restart Ollama.

💡 Multiple Origins Support

You can also allow local web apps like this:

OLLAMA_ORIGINS=chrome-extension://*,moz-extension://*,http://localhost:3000
🔍

4. Verify Ollama

Open this in your browser:

http://localhost:11434

Or use curl:

curl http://localhost:11434/api/tags
You should see a JSON response with available models.
⚙️

5. Configure Extension

1. Click the ⚙️ Settings icon in the extension popup.

2. You can configure:

  • Base URL (http://localhost:11434)
  • Default model (gemma:3b, llama3, etc.)
  • Theme and other preferences
🧯

Troubleshooting

Common Issues:

  • Ensure ollama serve is running
  • Make sure your model is pulled: ollama pull <model>
  • Avoid firewalls/VPNs that block localhost

Still Having Issues?

  • Verify your OLLAMA_ORIGINS is correctly set
  • Check browser console for error messages
  • Restart Ollama service after config changes
📎

Helpful Links

🌐 Documentation

Ollama Docs

🧩 Extension

Chrome Store

💻 Source Code

GitHub