Google I/O 2026 unveils Gemini 3.5 Flash and Omni AI models

Google unveiled Gemini 3.5 Flash as the default model for its Gemini app and AI‑powered Search, promising faster response times and stronger guardrails against harmful output [1]. The launch also introduced Gemini Omni, a multimodal model that can generate video from text, photos, audio or existing clips, debuting today in the Gemini app, Google Flow and YouTube Shorts [1].

Gemini 3.5 Flash arrives with a “neural expressive” redesign of the Gemini app, rolling out on the web and on Android and iOS starting May 19. Google claims the model handles long‑horizon agentic tasks at roughly half the cost of comparable frontier models, and it outperforms Gemini 3.1 Pro on benchmarks such as Terminal‑Bench 2.1 and GDPval‑AA [1][2]. The company also announced Gemini Spark, an always‑on AI agent that runs in the background on Google Cloud, linking to Workspace apps like Docs, Gmail and Sheets to automate tasks such as email drafting and expense monitoring [1].

Developers can now “vibe‑code” entire native Android apps within AI Studio, complete with an embedded emulator for testing and direct deployment to the Play Store. Google plans to let users publish these AI‑generated apps to a limited audience before a public release, and future Firebase integration will expand the ecosystem [1].

Beyond AI models, Google showed an updated Project Aura smart‑glasses prototype with a redesigned compute puck and fingerprint sensor, and announced two new audio‑only XR glasses from Warby Parker and Gentle Monster slated for fall release [1]. The company also rolled out a “Universal Cart” that aggregates products from multiple merchants across Search, YouTube and Gmail, and expanded AI‑driven search capabilities to accept text, images, files, videos and Chrome tabs as inputs [1].

These announcements signal Google’s push to embed generative AI deeper into everyday tools, positioning Gemini 3.5 Flash as the backbone for both consumer features and developer workflows. The real question now is how quickly developers and users will adopt the new AI‑first app creation flow, and whether Gemini Omni’s video generation will reshape content creation across Google’s platforms.