It’s 2026 and Google wants AI to be a part of almost everything that you do. Be it shopping, writing emails, making dinner reservations, planning for vacations, and pretty much anything that you can imagine. And that is exactly what Google I/O 2026 was all about. The event started livestreaming on YouTube at 10:30 PM (IST) on May 19. The very first thing that we got to hear were, “Take a second to realise how far we have come”.
In the next 2 hours, Google made announcements regarding AI, a new agentic era, and well, more on the AI side. Talking about hardware, we got to see a new pair of smart glasses that Google is developing in collaboration with Samsung.
Now, all of this sounds vast, and it really is. Without further ado, let’s take a look at everything that was announced last night at Google I/O 2026.
Also read: Google I/O 2026: Wear OS 7 launches with better battery life, Gemini AI and smarter features
As mentioned already, a large part of the keynote was dominated by announcements related to Gemini. Google has officially launched the Gemini 3.5 Flash, which is now the default Gemini model across several Google AI products. The company claims the model is faster, more efficient, and performs better than previous Gemini versions in most benchmarks. Google also confirmed that Gemini 3.5 Pro, a more powerful flagship model, will launch in June this year.
Now if you have been on the fence about buying a Gemini subscription because of its price, here’s some good news. Google announced pricing changes for its AI subscription plans and they cost way less than before.
AI Ultra now starts at $99.99 (Rs 9,685 approx) per month instead of the earlier $250 pricing. As for the previous $250 tier, it still exists but now costs $200 (approx Rs 19,370) monthly. Google’s updated AI plans now include AI Plus ($7.99), AI Pro ($19.99), and AI Ultra.
The company is also replacing prompt-based usage limits with compute-based limits, meaning heavier AI tasks like video generation will consume more quota than simple text prompts.
Gemini Spark is a new cloud-based AI agent that has been designed in a way that it will continuously work in the background and perform tasks on behalf of users.
Google’s Josh Woodward stepped on to the stage to introduce Gemini Spark and shared a few examples where it can be useful. He showed us how we can find the Gemini Spark button the moment we open Gemini and check all the background tasks it has been indulging in so far. Then, Woodward asked Spark to write an email to his team and plan a party, which it did in minutes.
The cool thing is that Spark can connect with Gmail, Docs, Uber, OpenTable, Lyft, Zillow, and several third-party services to manage such workflows.
Woodward ended his demo by saying that Spark will keep working in the background on all his given tasks, even as he puts his phone down.
In addition to this, to ensure users maintain control, the company also introduced Agent Payments Protocol (AP2), which lets users define spending limits, approved stores, and purchase restrictions for AI agents. As far as availability is concerned, Spark will initially roll out to AI Ultra subscribers in the US.
Another major announcement from Google was about Search, and how it will look completely different. Google unveiled a redesigned AI-powered Search Box that can understand natural language conversations while also accepting images, videos, files, and Chrome tabs as inputs.
The company is also making AI Overviews interactive. Thanks to this, you can now ask follow-up questions directly inside search results instead of getting static summaries. The intent here seems to be to turn Search into a conversational chatbot-like experience.
Google also announced Search Agents which are AI assistants that can continuously monitor websites, blogs, listings, and social media pages for updates. The company showcased examples like apartment hunting and sneaker drop tracking.
Google is also expanding Personal Intelligence in Search, allowing Gemini to pull context from Gmail, Photos, and Calendar for more personalised results.
Another addition worth mentioning is AI-powered calling. Google says Search agents will soon be able to call businesses on behalf of users to gather information or complete bookings.
Google didn’t forget its Gmail and Workspace users and announced several AI-focused features that have the potential to enhance productivity.
Let’s start with Gmail Live, which is a conversational voice-based assistant for inbox searches. You can verbally ask questions like flight timings, school schedules, or meeting details without manually searching through emails.
AI Inbox is also getting smarter with personalised draft replies, faster access to Docs and Sheets, and AI-powered task management tools.
Meanwhile, Docs Live can now transform rough notes and fragmented thoughts into structured documents in real time using Gemini AI.
Not just this, Google also announced Daily Brief, a feature that generates personalised summaries using Gmail, Calendar, and tasks. The feature is rolling out first in the US for AI Plus, Pro, and Ultra users.
Another new tool called Google Pics is all about AI image creation and editing. Google says users can create posters, social media visuals, flyers, and edited graphics through AI prompts.
Rumours of Google developing AI Smart Glasses have been around for months. And at Google I/O last night, we finally got to hear about them. Google’s Shahram Azadi presented the AI smart glasses, developed in collaboration with Samsung and Qualcomm. He also said that these are designed to deliver AI assistance without users having to constantly look at their phones.
Also, there are two types of these glasses. The first is audio glasses, which provide spoken responses and AI assistance through built-in speakers. The second category is display glasses, which can show information directly in front of the user’s eyes. Audio glasses will launch first later this fall.
To make the glasses feel more wearable and fashion-focused, Google has partnered with Warby Parker and Gentle Monster. The company showcased some early designs during the keynote and said more styles will arrive later this year.
The most important aspect is that the glasses are deeply integrated with Gemini. Users can activate Gemini either through voice commands or by tapping the frame. Once activated, the assistant can answer questions about surroundings, identify objects, provide restaurant information, explain signs, and even help with navigation.
Now most of us spend time shopping online. And that too is going to be powered by AI.
A new Universal Cart feature combines products added across Search, Gemini, Gmail, and YouTube into one central shopping hub powered by Google Wallet.
Google also introduced Universal Commerce Protocol (UCP), which allows AI agents to complete purchases and bookings directly through partners including Amazon, Walmart, Shopify, and Meta.
Google also announced Gemini Omni, a new multimodal AI model designed for advanced video generation and editing.
Unlike standard text-to-video tools, Omni can accept text, images, audio, and videos as inputs and generate content accordingly. The company demonstrated conversational editing tools capable of changing backgrounds, adding effects, adjusting scenes, and even generating digital avatars.
Google also updated its AI filmmaking platform Flow with Gemini Omni Flash integration. Users can now perform multi-step cinematic edits, create music, add characters, and modify scenes through prompts.
All AI-generated videos created through Omni will include Google’s SynthID watermarking technology.
Apart from all this, three other announcements were made. First one is that Google revealed how SynthID, its AI watermarking technology, is now being adopted by companies including OpenAI, ElevenLabs, and Kakao.
The company says Chrome and Search users will soon be able to identify AI-generated content by checking for SynthID markers directly inside browsers and search results.
On the other hand, YouTube also received several AI-focused announcements. For instance, Gemini Omni is coming to YouTube Shorts Remix, allowing creators to generate and edit AI-enhanced Shorts using conversational prompts. Google is also expanding its likeness detection tools to help creators identify AI-manipulated versions of their faces and videos.
Another feature called Ask YouTube will allow Google Search users to surface relevant tutorial videos directly inside search results for complex questions.
And finally, Google previewed CodeMender, an AI-powered security tool capable of automatically detecting and fixing software vulnerabilities. The feature is currently being tested with select security researchers ahead of a wider rollout.
After looking at all these announcements, one thing is clear – the tech giant no longer wants AI to just be an app or something you turn to occasionally. It wants people to rely on AI for even the most basic things that they do online like shopping, writing emails, planning vacations, creating videos and much more.
Almost every announcement at Google I/O 2026 revolved around reducing the need to manually interact with devices.
Now as Gemini becomes more deeply integrated into Search, Gmail, shopping, and even wearables, users may eventually have to decide how much control they are comfortable giving AI systems. The technology showcased at I/O feels futuristic and useful in many cases, but it also points to a future where AI could become the main layer between people and the very devices they own as well as the internet. The question is – Are we really ready for that?