source: google ai: i/o 2026
level: technical
google io 2026 focused on making ai more helpful through new models and agentic experiences. two new models were announced: gemini omni and gemini 3.5 flash. gemini omni can create anything from any input, starting with video, and offers improved world understanding, multimodality, and editing. gemini 3.5 flash is the first in a new family combining frontier intelligence with action, built for complex agentic workflows.
the company highlighted google antigravity, an agent-first development platform that moves beyond writing assistance to agents that help users act. these agents enable anyone to be a builder. agentic features are rolling out across products, including information agents in search, gemini spark and daily brief in the gemini app, and universal cart, an intelligent shopping cart. other updates include intelligent eyewear for android, new voice capabilities in workspace, and google pics for design.
developer tools saw updates to google antigravity, an enhanced gemini api, and native android support in google ai studio. google also introduced gemini for science, a collection of tools for scientific exploration, and co-scientist, a multi-agent ai partner for research. digital watermarking expanded with synthid and c2pa content credentials to identify ai-generated media. new subscription plans include a $100 ai ultra plan.
why it matters: these releases push ai from passive assistance to proactive agents that can act across apps, changing how developers build and users interact with technology.
source: google ai: i/o 2026