Sunday, December 22, 2024

What are Gemini Extensions? Make Google’s chatbot smarter than ChatGPT

Must read

Rita El Khoury / Android Authority

If you rely on your phone’s digital assistant for automation or even just hands-free control, you may have hesitated transitioning from the Google Assistant to Gemini. The latter developed a reputation for lacking many of the Assistant’s features early on. However, Google will only support Gemini going forward and has even made it as the default assistant on the Pixel 9 series. Luckily, Google is bridging the gap between the Assistant and Gemini with “extensions” for Gemini that allow the chatbot to perform useful real-world tasks.

With Gemini Extensions enabled, the chatbot can do everything from managing your to-do list to creating a music playlist based on your current mood. And in the near future, credible leaks indicate that new extensions will allow you to send WhatsApp messages and control your smart home. So with that in mind, here’s what Gemini Extensions can do and which ones you should consider using to make your life easier.

What are Gemini Extensions?

As the name suggests, Gemini Extensions give the AI chatbot access to various web services such as Google Maps, YouTube Music, and Gmail.  And while the official list is currently limited to Google’s first-party services, we have also heard rumors of Spotify and WhatsApp extensions in the works.

But why would you want an AI to interact with third-party services or even Google’s own web apps? Simply put, extensions allow Gemini to access personalized or real-time information. This results in accurate and grounded responses. Remember when ChatGPT was criticized for hallucinating or responding with false information? Gemini avoids this by accessing information from external sources depending on your query.

Extensions allow Gemini to fetch information from the real world…and your Google account.

For example, Gemini can consult the Google Flights, Maps, and Hotels extensions to piece together a trip itinerary that includes actual travel times, distances, and costs. Likewise, I have personally used Gemini to look for long-lost information buried within obscure emails. With the right extension enabled, you can ask an open-ended question like “How much did my hotel in Tokyo cost me last year?” and Gemini will reply with an amount within seconds. This saves me the trouble of opening the Gmail app, searching for the right email, and scanning through the booking confirmation.

That aside, the feature comes in especially handy when you need to summarize long YouTube videos. Whenever I come across an interesting recipe video without a written version, I simply paste the YouTube URL into a new Gemini prompt and ask for a summary. On Android, the Gemini mobile app has deeper integration with the YouTube app. Simply tapping the “Ask about this video” chip (pictured above) allows you to ask follow-up questions without leaving the YouTube app.

The best part about Gemini Extensions is that everyone can use it for free. Unlike many of the chatbot’s newer features like Gems and voice chat, you don’t need to shell out $20 monthly for a Gemini Advanced subscription.

How to enable and use Gemini Extensions

Each Gemini Extension can be toggled on or off individually, which allows you to prevent the chatbot from reading your personal data. As you would expect then, the Google Workspace and YouTube Music extensions remain disabled by default — understandable as they grant Gemini access to your emails, documents, and playlists. You can navigate to Settings > Extensions within the Gemini app on desktop or Android and enable them manually. Most other Gemini Extensions like Flights and YouTube should already be enabled.

Using Gemini Extensions is straightforward — you do not need to specify a particular service in your prompt. A prompt like “Show me flights to Tokyo departing December 3 and give me ideas of things to do for a five day trip” will automatically invoke the Flights and Maps extensions. Gemini will also reference your current location to search flights from your nearest airport. If you’re happy with the results, you can then output it to a spreadsheet with a prompt like “Create a spreadsheet itinerary for the five days”.

The best Gemini Extensions and how to use them

Gemini Ask This Page Early Hands On

C. Scott Brown / Android Authority

With the basics of Gemini Extensions out of the way, which ones should you actually use? The list is a bit sparse at the moment since you can only choose from Google’s first-party services like Workspace and YouTube. Still, here is our ranking of the best Gemini Extensions and when you should use them:

  • YouTube: With the YouTube extension enabled, Gemini can save you from watching long videos if you’re short on time. And if you use the Gemini Android app, the “Ask about this video” feature makes it even more convenient. Whether you need a quick text summary or a step-by-step written guide, the YouTube extension can do it all and gets our top recommendation.
  • Google Workspace: Given that most of us live within Google’s ecosystem, the Workspace extension is useful if you have thousands of emails and documents. Once enabled, the chatbot can answer questions like “How much did I spend on Uber trips between January and May 2024?” Likewise, you can ask it to summarize long documents — Gemini Advanced shines here with its significantly larger context window that allows it to handle 1,500 page-documents at a time.
  • Flights and Hotels: Most other chatbots cannot help you plan a trip as thoroughly as Gemini can with its Google Flights and Hotels extensions enabled. I typically use both of these services on my desktop, but the chatbot interface allows me to quickly specify conditions like “I need a daytime flight” or “hotels under $100”.
  • Maps: The Maps extension may seem boring at first glance — you can use it to ask Gemini for directions to your nearest Starbucks but it doesn’t stop there. With a prompt like “Suggest the best places to get claypot rice in Hong Kong based on reviews,” you can save yourself several minutes of research. The chatbot even embeds a map at the end of its response.
  • YouTube Music: A while ago, I wrote about how Google could infuse its music streaming service with generative AI to create a more personalized DJ experience. Unfortunately, the YouTube Music extension for Gemini doesn’t realize that dream. It can search for songs, artists, and playlists but I could not get it to generate brand new playlists or suggest artists based on my listening habits.

These are the only Gemini Extensions currently available, but we have spotted references to upcoming extensions in the Android app that will allow Gemini to place phone calls, dim your smart lights, and read messages from chat apps. We will keep this list updated with new extensions and their use cases as Google decides to release them.

Latest article