From Gemini Live to Deep Think, Google makes its AI smarter and more integrated with updates and new features updated

It looks like Google has made some interesting updates and added some intriguing new features to Gemini.
#google #googleio2025 #ai #gemini

By Ken Wong - 24 May 2025

Note: This article was first published on 21 May 2025.

*Updated: 8 June 2025 with Veo3 local release and Google AI Ultra subscription information*

Some of Google's Gemini highlights. Photos: Google.

At Google I/O 2025, following the trend of past years, it's all about AI. This year, one of Google’s big announcements is how its Gemini app is getting even smarter with new features and integration into other apps and products.

These updates come at a time when Singapore is pushing the boundaries of AI usage and adoption. According to aitools.xyz March 2025 saw Singapore ranked 34th amongst the countries in total traffic to AI tools, with 76.71 million visits and a 0.61% global share. This was a monthly increase of +14.38 million visits, signalling growing adoption and rising public and institutional interest in AI technologies.

In fact, when adjusted for population size, aitools.xyz reported that despite on;ly having a population of around 6 million, Singapore has 12.78 AI tool visits per person. This is nearly double Australia’s (7.42) and Canada’s (7.38), and even ahead of the United States (6.07).

AI access doesn’t come cheap

As we covered earlier, Gemini has received a number of updates covering improvements to existing models to a host of new and interesting features. While access to most of them are free, some do require a subscription. Starting today, Google is introducing two subscriptions: Google Al Pro and Google Al Ultra.

With Google AI Pro, you'll get access to a suite of AI tools for S$28.99/month. This Pro plan will level up your Gemini app experience and replace and expand on Gemini Advanced. It also includes products like Flow, NotebookLM and more, all with special features and higher rate limits.

The second subscription is called the Google AI Ultra plan. It provides access to Google’s most powerful models with the highest rate limits and early access to any experimental Al products before anyone else. Unfortunately, Google AI Ultra is available in the U.S. only for $249.99/month, with an offer for first-time users of 50% off for their first three months. Google says it will be available for more countries coming soon.

Some of the updates, new, and upcoming features include:

Gemini Live

The update to Gemini Live allows camera and screen sharing to be available to everyone on Android and iOS for free, starting today. What this means is that you can point your phone at anything you may need help with, or ask it questions about something on your screen, and it will answer you in voice mode.

Imagen 4 and Veo 3

Imagen 4 is Google’s new AI image generation that delivers visuals that are clear and sharp in lifelike and fine details like intricate fabrics, water droplets, and animal fur, and better text and typography outputs, while suiting both photorealistic and abstract styles. Everyone can try Imagen 4, today in the Gemini app.

*Updated*

Veo 3 is a video generation model that Google says is the first in the world to have native support for sound effects, background noises and dialogue between characters meaning you can use text and image prompting matched to real-world physics for accurate lip syncing. It is able to create video with greater realism and fidelity, including 4k output based on real world physics and more accurately match your creative vision and instructions thanks to improved prompt adherence.

For example the video above was created in Veo3 with the prompt:

Prompt: A medium shot frames an old sailor, his knitted blue sailor hat casting a shadow over his eyes, a thick grey beard obscuring his chin. He holds his pipe in one hand, gesturing with it towards the churning, grey sea beyond the ship's railing. "This ocean, it's a force, a wild, untamed might. And she commands your awe, with every breaking light"

Veo 3 is available today for Google AI Ultra subscribers in 73 countries including Singapore, while Google AI Pro subscribers get a limited number of tries over a limited period.

Deep Research and Canvas

Available today, Deep Research creates customised reports by trawling the web to find content from online websites and combines the data it finds with uploaded content from your own files and records.

You can then turn the results over to Canvas and make your reports even more interesting by turning them into interactive content, quizzes, Audio Overviews (with Canvas being available in 45 languages), and more.

Gemini in Chrome

Gemini in Chrome will begin rolling out to users who have English as their Chrome language on Windows and macOS devices. Google says that this first version allows you to easily ask Gemini to clarify complex information on any webpage you're reading or to summarise the information. In the future, Gemini will be able to work across multiple tabs and navigate websites on your behalf. Unfortunately, it's currently only available for Google AI Pro and Google AI Ultra subscribers in the U.S.

Gemini Quizzes

Gemini can change the way you study by creating interactive quizzes designed to make learning more engaging. Google gives the example of asking Gemini to “create a practice quiz on thermodynamics”. Once created, you have a tailored learning experience that provides instant feedback, highlighting topics that need more attention as you answer. Even better, once you're done, Gemini will proactively offer a personalised follow-up quiz, focused on the areas you found challenging, helping you turn weaknesses into strengths. This feature is rolling out today to all Gemini users around the world on desktop and mobile.

Deep Think

Google is starting to test an enhanced reasoning mode called Deep Think for Gemini 2.5 Pro, that will allow Gemini to essentially “think harder” about your questions and not just respond with the first answer as quickly as it can. This will allow it to consider multiple possible hypotheses before answering you. It is currently available to trusted testers via the Gemini API to get their feedback before wider available.

Improvements to Flash

Improvements to Flash have been made, improving its speed and efficiency, using fewer tokens while scoring higher in benchmarks for reasoning, multimodality, code and long context than its predecessor. For now, the new 2.5 Flash is now available for preview in Google AI Studio for developers, in Vertex AI for enterprise, and in the Gemini app for everyone.

Our articles may contain affiliate links. If you buy through these links, we may earn a small commission.