Google turns Gemini right into a proactive AI agent with Spark, Day-to-day Transient, and main redesign

Google simply introduced one of the vital greatest overhauls to the Gemini app but at Google I/O 2026, introducing a redesigned interface, proactive AI brokers, and a strong new video era fashion referred to as Gemini Omni.

The corporate claims Gemini is now utilized by greater than 900 million folks throughout 230 nations and territories, making it “probably the most broadly to be had generative AI device on this planet.” As a part of its newest push to make Gemini much more tough, Google is now evolving the chatbot right into a extra proactive assistant that may actively lend a hand organize your virtual existence.

The most important visual alternate to the Gemini app is a brand new design language Google calls “Neural Expressive.” The refreshed interface is already rolling out to customers and introduces fluid animations, up to date typography, colourful colours, haptic comments, and a extra dynamic presentation taste.

Google says Gemini will now generate responses with richer formatting, together with pictures, summaries, bolded textual content, interactive graphics, timelines, or even narrated movies to keep away from appearing customers massive partitions of textual content.

Google says it’s also absolutely integrating Gemini Reside into the core Gemini revel in. Customers can now transfer seamlessly between typing and herbal voice conversations with out breaking context. The corporate says it additionally redesigned the microphone revel in so customers can talk naturally with out interruption, and that reinforce for regional dialects is coming quickly.

The Gemini app’s Neural Expressive redesign begins rolling out globally nowadays throughout Android, iOS, and the internet.

At I/O 2026, Google additionally offered Gemini Omni, a brand new multimodal fashion serious about video introduction. Omni has been doing the rounds on the net in recent times, and will mix textual content, pictures, and video activates to generate top of the range cinematic movies. Google says customers will have the ability to get right of entry to Omni immediately from the Gemini app.

Customers can add pictures from their digicam roll and edit it the usage of herbal language activates fairly than conventional enhancing equipment. Google additionally says customers will have the ability to create AI avatars that glance and sound like them.

Gemini Omni may be rolling out nowadays to Google AI Plus, Professional, and Extremely subscribers international.

Past the Neural Expressive redesign and Omni, Google’s large upgrades to Gemini additionally come with two logo new AI brokers — Day-to-day Transient and Gemini Spark.

Because the identify suggests, Day-to-day Transient is a customized digest that summarizes your day the usage of data out of your calendar, reminders, go back and forth plans, and extra. It’s a characteristic Google up to now tried to release as Day-to-day Hub at the Pixel 10 collection, however got rid of because of deficient person comments.

Google describes Gemini’s Day-to-day Transient as a morning briefing revel in that prepares you for the day forward. It’s launching nowadays throughout the Gemini app for Google AI Plus, Professional, and Extremely subscribers in the USA.

The second one, and extra formidable, characteristic is Gemini Spark. Google describes Spark as a 24/7 cloud-based AI agent that may proactively maintain duties for your behalf even after you shut your computer or lock your telephone.

Powered by way of Gemini 3.5 Flash, Spark deeply integrates with Workspace apps like Gmail, Medical doctors, and Slides. Google shared examples like parsing bank card statements to spot hidden subscriptions, tracking faculty emails for points in time, growing day by day digests for households, or even robotically turning scattered assembly notes into polished Medical doctors and draft emails.

Spark too can create routine workflows and triggers. Google says customers stay in keep watch over, and Spark will ask for permission earlier than taking high-stakes movements like spending cash. Talking of which, Google has created a brand new Agent Bills Protocol (AP2) to lend a hand brokers like Spark make protected bills for your behalf. The corporate says it’ll get started bringing AP2 to Google merchandise within the coming months, beginning with Spark.

Gemini Spark rolls out to relied on testers this week, with a beta release for US Google AI Extremely subscribers anticipated subsequent week.

Google may be increasing Gemini’s attached app ecosystem thru new integrations launching nowadays with Canva, OpenTable, and Instacart. Spark will ultimately use the ones integrations to finish duties throughout those products and services. Further features deliberate for later this yr come with Chrome integration, texting and emailing Spark at once, customized sub-agents, and native browser keep watch over.

In the end, Google showed that its macOS Gemini app is getting deeper Spark integration and new voice features, permitting Gemini to paintings extra at once with native system workflows.