The writing has been on the wall for a while, but the shift away from Google Assistant is now official. Google has announced that it will shift users to Gemini as the default AI assistant on their devices in the coming months. Once that happens, they will no longer be able to access the Google Assistant.

At the moment, you can switch to Google Assistant as the default option on your Android phone, even on newer phones that come with Gemini running out of the box. In addition to phones, Google will be giving a similar treatment to smartwatches, Android Auto, tablets, smart home devices, TVs, and audio gear.

“We’re also bringing a new experience, powered by Gemini, to home devices like speakers, displays, and TVs,” says Google, without sharing a specific time frame for the transition. What happens to Google Assistant following the transition? Well, it will be removed from devices and will no longer be available to download from app stores.

Talking about apps, Gemini can already interact with a wide range of Google’s own as well as a few third-party apps. Users can ask it to perform chores across different products, without ever having to open those apps. In addition to in-house apps such as Docs, Drive, and Gmail, the Gemini assistant can also perform tasks in third-party apps such as WhatsApp and Spotify, alongside a bunch of Samsung apps.

Please enable Javascript to view this content

What devices are moving to Gemini?

Gemini will support a majority of the features that are currently available with Google Assistant. The fundamental Google Assistant capabilities, such as controlling smart home devices, accessing the virtual assistant from the Lock Screen, routines, song detection, and messaging will be available with Gemini.

A healthy bunch of them have also been enhanced for Gemini to offer a better experience. For example, Gemini can better handle tasks such as dealing with Reminders, events, and lists, tackling queries related to Google Maps, Google Drive, enquiring about YouTube videos, awareness of on-screen context, and more.

Following is the broad list of comparability for devices that will be upgraded to Gemini in the coming months:

• Phones with at least 2 GB RAM.
• Android 10 or a later version.
• iPhone running iOS 16 or a newer version.

The virtual assistant on your headphones will match the one currently installed on your phone. An exception will be the budget-centric Android Go phones, which run a simplified and pure Google-only experience of the operating system. No brand has launched an Android Go phone in years, and the whole project is effectively dead, so it’s not surprising that Gemini is not moving ahead for these devices.

How are things changing with Gemini?

Let’s start at the basic level. In case you are concerned about any change in the activation hotword, there is no reason for concern here. The “Hey Google” vocal cue will continue to work as the hotword for summoning Gemini, just the way it worked for Google Assistant. For devices that haven’t been transitioned to Gemini, using the “Hey Google” hotword will summon the Google Assistant without any feature limitations.

In order to make the switch seamless, especially in terms of familiarity for users, Google will also be porting over some of the user data from Google Assistant to Gemini. Assuming users have already provided the necessary on-device permissions, Gemini will automatically get access to details such as calls and messaging history.

“We will use some preferences and history from Google Assistant when we upgrade your device to Gemini,” says Google in a support document. Thankfully, users won’t be running into a functional wall. In the coming months, Gemini will also take a look at your Search history to give more personalized answers. Down the road, it will be able to interact with Photos, too.

Why is this a better deal?

Gemini is almost a fundamental upgrade over Google Assistant, down to the foundation tech stack. Where Google Assistant was an AI product built atop natural language understanding capabilities, Gemini stands out as a natural language chatbot that is capable of language understanding as well as reasoning.

Unlike Google Assistant, Gemini is more like a knowledge bank that has been trained on a whole corpus of information and works offline for a wide range of on-device tasks. It can handle free-flowing conversations in multiple languages, but where it reigns supreme is its understanding of various input formats.

For example, you can upload a PDF file and Gemini will break it down for you. It can accept pictures, text, audio, and even live video as an input format. In technical terms, it is called multi-modal processing. Gemini is also capable of image generation.

Talking about comprehension capabilities, Gemini can solve complex science, maths, and coding-related problems — the kind of stuff Google Assistant can’t handle. In the coming months, with the rollout of Project Astra, Gemini will also be able to understand the world around you, as seen through the camera in real-time.

But there are a few other features that go far beyond what Google Assistant could ever accomplish. For example, the newly free Deep Research tool can perform comprehensive web research and create a full-fledged report in a matter of minutes, a task that would otherwise take hours of internet search and documentation.

Then we have Gems, which is essentially a user-generated custom version of the Gemini assistant, trained to perform a certain task. Any person can create these Gems, without any coding knowledge, and by using only natural language instructions. So far, Gems and Deep Research have been limited to paying subscribers, but Google made them available to all users for free in March this year.

Another notable aspect of Gemini is that it comes in multiple formats, each with its own set of capabilities. Google Assistant was more of a one-size-fits-all approach. With Gemini, users can pick between models based on the complexity of the task at hand.

For example, if you want it to answer questions based on a 200-page book you uploaded, the “Pro” model is the best choice. Otherwise, the “Flash” model is snappy and capable enough for your average virtual assistant tasks.

In a nutshell, Gemini is a massive upgrade over Google Assistant, both in terms of sheer capabilities as well as versatility. And if you own a high-end phone that can run the Gemini Nano model natively, the AI assistant will handle tasks like live translation without even requiring an internet connection.