Google launches Gemini 3.5 Flash

Google used its May 19 I/O conference to tie together three pieces of its AI strategy that had been moving on separate tracks: a faster model, a more persistent agent, and new hardware designed around Gemini. The immediate product launch was Gemini 3.5 Flash, which Google said is available now in the Gemini app, AI Mode in Search, Google AI Studio, Android Studio and its enterprise tools. The companion software push was Gemini Spark, a new agent Google described as running in the background to help manage tasks across a user’s digital life. The hardware message was that Gemini is no longer just a feature inside phones and browsers, but the organizing layer for laptops, glasses and other devices. ### Why did Google make Flash the centerpiece? Google said Gemini 3.5 Flash is the first release in its new Gemini 3.5 family and positioned it as the model that combines “frontier intelligence with action.” In practice, that means Google is using Flash as the model it wants people to touch first, rather than keeping its most capable work behind a slower premium tier. (blog.google) Google said 3.5 Flash is its “strongest agentic and coding model yet,” outperforming Gemini 3.1 Pro on benchmarks including Terminal-Bench 2.1, GDPval-AA and MCP Atlas, while also leading on multimodal understanding in CharXiv Reasoning. The company also said the model produces output tokens at four times the speed of other frontier models. Those claims matter because Google is pitching low latency as part of model quality, not just as an engineering detail. (blog.google) ### What exactly is Gemini Spark supposed to do? Josh Woodward, Google’s vice president for Google Labs, Gemini app and AI Studio, said Gemini Spark is “a 24/7 personal AI agent” designed to proactively manage tasks under a user’s direction. Google paired that with a new Daily Brief feature, which it described as a personalized morning update, and with a redesigned Gemini interface that can move between typed prompts and live conversation. (blog.google) Google also said the macOS Gemini app will integrate Gemini Spark so it can operate on a local machine, alongside new voice features. That detail is important because it shows Google trying to move beyond the chatbot window and into software that can watch context, stay present and act across sessions. ### Why are laptops and glasses part of the same announcement? (blog.google) Google spent the week framing Gemini as a full-stack system, not just a model release. In his keynote, Chief Executive Sundar Pichai said Google was taking a “full-stack approach” spanning custom silicon, models, products and platforms. On May 12, Google introduced Googlebook as a new laptop category “designed for Gemini Intelligence,” with features such as Magic Pointer, custom Gemini-built widgets and closer integration with Android phones. (blog.google) Google said those devices will launch later this fall. On May 19, Google also said audio smart glasses built with Samsung and eyewear brands Gentle Monster and Warby Parker will arrive later this fall. (blog.google) Google said the glasses will support spoken help, directions, texting, photos, translations and multi-step task handling through Gemini. ### What is Google trying to prove with this rollout? (blog.google) Google said more than 900 million people now use Gemini monthly across 230 countries and more than 70 languages. Pichai also said Google is processing more than 3.2 quadrillion tokens per month across its surfaces and that more than 8.5 million developers are building apps and experiences with its models monthly. (blog.google) Those figures suggest Google wanted I/O 2026 to show that Gemini is no longer an experimental sidecar. The company’s own framing was that AI products are moving from tools that help people write to agents that help them act. ### What should readers watch next? Google said Gemini 3.5 Pro is expected next month, giving the company a second chance to define the top end of the family after Flash’s launch. (blog.google) Googlebook updates are due later this year on Google’s product pages, and the first Gemini-powered audio glasses are scheduled for later this fall through Samsung, Gentle Monster and Warby Parker. (blog.google 1) (blog.google 2)