Models

Google I/O 2026: Gemini 3.5 Flash and Spark Agent Land in Gemini App

| By AI Revolution Editorial Team
Aerial photograph of a hyperscale data-center campus at dusk, illustrating the infrastructure behind Google's Gemini 3.5 Flash model and the Gemini Spark personal agent unveiled at Google I/O 2026

Google opened I/O 2026 on May 19, 2026 with a Gemini-heavy keynote that centered on three product moves: a new default model in the form of Gemini 3.5 Flash, a personal agent called Gemini Spark, and a new multimodal series branded Gemini Omni. According to Google's blog and live coverage from 9to5Google and MacRumors, 3.5 Flash is rolling out the same day to the Gemini app, AI Mode in Search, Antigravity 2.0, and the Gemini API.

On internal benchmarks Google says 3.5 Flash surpasses last year's Gemini 3.1 Pro on coding and agentic tests — 76.2% on Terminal-Bench 2.1, an 1,656 Elo on GDPval-AA, 83.6% on MCP Atlas — and on the multimodal CharXiv test at 84.2%. The pitch is intelligence rivaling last-generation flagships at Flash-tier latency and roughly half the cost of comparable frontier models. A larger Gemini 3.5 Pro is in testing and is expected next month.

The headline consumer story is Gemini Spark, a 24/7 personal agent that runs on 3.5 Flash and is positioned as a shift from an assistant that answers questions to one that takes actions on your behalf. Spark is rolling out first to trusted testers and reaches Google AI Ultra subscribers in the US in beta next week. Around it, Google introduced Daily Brief, which sifts through Gmail, Calendar, and Tasks to compose a personalized digest, and Gemini Omni, a new model series that accepts image, audio, video, and text input and generates video grounded in real-world knowledge. The Gemini app also picked up a new design language called Neural Expressive, with fluid animations, fresh typography, and haptic feedback.

Sources

Google Blog (blog.google), 9to5Google, MacRumors, Google Cloud Blog

AI Revolution Editorial Team

Weekly AI intelligence by SBN Media Studio. Learn more about our team.

← Back to Models