Google unveils new Gemini models, AI Glasses, and Agentic Tools at I/O 2026

Google unveiled AI tools at I/O 2026, including Gemini Omni, AI Search, smart glasses, and Antigravity 2.0. Here's everything announced.

Google I/O 2026

More tools to play around with

Google announced a range of artificial intelligence tools and updates at its annual I/O developer conference in 2026, spanning models, search, hardware, and productivity applications.

New Gemini models

1. Gemini Omni - This feature is a multimodal model designed to generate media outputs from combinations of text, photos, and video clips. The model is launching first through the Gemini app, Flow, and YouTube.

2. Gemini 3.5 Flash - Google's latest model focused on agentic and coding tasks. Google says it offers frontier-level performance at faster speeds and at a lower cost compared to competing models.

3. New AI Search Box - Google is redesigning its Search interface for the first time in over 25 years. The updated search box accepts text, images, files, videos, and Chrome tabs as inputs, with an interface that expands dynamically based on the query.

4. Intelligent Eyewear Google - The newly announced audio glasses are powered by Android XR, scheduled to launch this fall. The glasses are designed for hands-free use and deliver Gemini-powered assistance through a private audio channel, while also supporting music playback, photography, calls, and access to phone apps.

5. Daily Brief - This is a new agent within the Gemini app that aggregates a user's daily digital information into a single morning summary, along with suggested next steps.

6. Gemini Spark - It is an always-on personal agent within the Gemini app that can take actions on a user's behalf, such as monitoring and flagging new subscriptions from credit card statements.

7. Neural Expressive - Google updated the Gemini app's interface under the name Neural Expressive, introducing new animations, typography, colors, and haptic feedback.

8. Google Antigravity 2.0 - It's a new standalone desktop application that allows users to run and coordinate multiple AI agents simultaneously. Use cases cited include parallel execution of tasks such as coding and asset generation.

For more information, click each link or head to the official Google tech blog page.

What do you guys think?