At I/O 2026, Google presented a set of generative and agent-driven products that sketch the next wave of AI-enabled user tools.

New foundational models

Google unveiled Gemini 3.5 Flash and Gemini 3.5 Pro. Flash is available now and, according to the company, improves speed, coding and agent capabilities.

Multimodal generation

The company introduced Gemini Omni, a multimodal model able to generate and edit diverse content formats, including images and video using a user's likeness.

Always-on personal agent

Google showcased Gemini Spark, a background agent designed to run continuously, integrate with Gmail, Calendar and Docs, and connect to third-party services.

The agent is described as capable of executing tasks autonomously and continuing workflows even after a device is closed.

Wearables and interaction

Google revealed smart glasses powered by Gemini that offer real-time translation, voice control, navigation and direct integration with the assistant for hands-free use.

Image editing with Nano Banana

Google Pics is a new image editor based on the Nano Banana model, enabling rapid object selection and edits with tools positioned as comparable to traditional desktop editors.

AI agents inside Search

Google Search will be able to run multiple AI agents concurrently for tasks such as price monitoring, release tracking, stock surveillance and product drop notifications.

Subscription changes

Google introduced a new Ultra subscription tier priced at $100 per month for access to enhanced AI features across its services.

The announcements emphasize always-on agents, expanded multimodal capabilities and deeper integration across Google's ecosystem, framing the company's immediate AI product roadmap.