This guide shows you how to run Google Assistant on Windows, Mac, or Linux desktops. While Google doesn’t offer an official desktop app, you can use the Assistant in your browser or via an open-source project that taps into Google’s own APIs. You’ll need a Google account and, for the API method, a bit of patience with developer tools.
What You’ll Need
- A Google account (Gmail works)
- A desktop computer running Windows 10/11, macOS 11+, or a current Linux distribution
- For the browser method: Chrome, Edge, Firefox, or Safari
- For the API method: Python 3.7 or newer, a code editor, and about 20 minutes
Step-by-Step
Method 1: Use Google Assistant in Your Browser (All Platforms)
- Open your web browser and go to google.com.
- Make sure you’re signed in to your Google account (click “Sign in” in the top-right if not).
- Click the mic icon in the search bar. If it’s your first time, your browser will ask for microphone permission – click “Allow.”
- Speak your query, or type it after clicking the mic. The results page will show a Google Assistant-style response card with a voice readout.
- To get a more assistant-like experience, install the Google Assistant extension from the Chrome Web Store (works on Edge and Brave too). This adds a dedicated button to your toolbar.
- Click the extension icon, grant microphone access, and you can now say “Hey Google” or click the button to ask questions, set timers, and control smart home devices.
Method 2: Use Google Assistant for Desktop (Open-Source App)
This method installs a standalone app that uses Google’s official Assistant API. It works on all three operating systems.
- Install Python 3.7 or later from python.org (Windows/Mac) or your package manager (Linux:
sudo apt install python3). - Open a terminal (Command Prompt on Windows, Terminal on Mac/Linux) and install the required library:
pip install google-assistant-sdk[samples] - Go to the Google Assistant SDK page and click “Get started.”
- Follow Google’s instructions to create a project in the Actions Console. You’ll need to enable the Assistant API and create OAuth credentials. This is the most technical step – take your time.
- Download the
client_secret.jsonfile from the credentials page and save it to your computer. - In your terminal, run:
python -m google.assistant.sdk.sample_audio(Windows) orpython3 -m google.assistant.sdk.sample_audio(Mac/Linux). - Follow the on-screen link to authorize your Google account. Copy the code it gives you, paste it back into the terminal, and press Enter.
- Once authorized, you’ll see a message like “Press Enter to send a request.” Press Enter, then speak. The Assistant will respond via your speakers.
- To make it a proper app, download the Google Assistant for Desktop wrapper from GitHub (search for “google-assistant-desktop”). It bundles the API setup into a friendly interface.
Method 3: Use a Third-Party Client (Mac Only)
- If you’re on macOS, download Assistant for Google from the Mac App Store (free).
- Open the app, sign in with your Google account, and grant microphone access when prompted.
- Click the menu bar icon and start talking. It uses the same API as Method 2 but with a native Mac interface.
Common Mistakes
- Microphone permissions not granted – The browser or app can’t hear you if you clicked “Block” earlier. Check your browser’s site settings or system privacy settings.
- Using the wrong Google account – The Assistant is tied to the account you authorize. If you’re signed into multiple accounts, make sure the right one is active.
- Skipping the OAuth step – In Method 2, if you don’t complete the authorization flow, the app won’t work. The terminal will hang or show an error.
- Python not in PATH – On Windows, if
pythonisn’t recognized, usepyinstead, or reinstall Python and check “Add to PATH.” - Expecting full Assistant features – The API version doesn’t support every feature (e.g., routines, some smart home commands). It’s a basic voice assistant.
Tips
- For the best experience, use a good USB or 3.5mm microphone – laptop mics pick up background noise.
- Set a keyboard shortcut for the desktop app (e.g., Ctrl+Alt+A) so you can invoke it without clicking.
- If you use multiple devices, link your Google Home devices to the same account, and the Assistant on your desktop can control them.
- For privacy, review your Google Activity settings and delete voice recordings regularly.
- Keep the Python library updated:
pip install --upgrade google-assistant-sdk
Troubleshooting
- “Device model not found” error – In the Actions Console, make sure you’ve registered a device model (under “Device registration”). Give it any name, e.g., “my-desktop.”
- Microphone not working in the browser – Click the lock icon next to the URL, go to Site Settings, and set Microphone to “Allow.” Then reload the page.
- Assistant responds but doesn’t speak – Check your system volume and make sure the app isn’t muted. On Linux, you may need to install
pulseaudio-utils. - Authorization loop – Delete the
credentials.jsonfile in the app’s directory and re-run the setup. - App crashes on startup – Ensure Python and all dependencies are installed correctly. Try running the sample script directly to see the error message.
- No internet connection – The Assistant requires an active connection. Check your firewall – some security software blocks the API ports.
Frequently Asked Questions
Can I use Google Assistant offline?
No. Both the browser and API methods require an internet connection to process your requests.
Is this the same as Google Assistant on my phone?
It’s close, but not identical. The desktop versions lack some phone-specific features like calling, and the API method may not support every smart home command.
Do I need to pay for anything?
No. The Assistant API is free for personal use, and all the tools mentioned are open-source or free.
Will this work on a Chromebook?
Yes, but Chromebooks already have built-in Assistant. You can use the browser method if you prefer.
Can I change the wake word from “Hey Google”?
No. The wake word is fixed. You can only trigger it by saying “Hey Google” or clicking the button.
Is my voice recording stored?
Yes, if you’re signed in. You can view and delete recordings in your Google Activity page.
Related
General steps for current mainstream versions; menus can differ by device and version. See the disclaimer or report an error.
More in this bed
- Fix msmpeg2vdec.dll Errors in Windows
- How to Update Sonos App and Speakers
- Unlock iPad Without Apple ID: Is It Possible?
- iOS 18 WhatsApp Notifications Not Working? Fix
- Fix Windows Cannot Locate Suitable Printer Driver
- Can You Check Who Viewed Your Google Photos?