Google rolls out Gemini 3.5 Flash

Google used the opening day of I/O on Tuesday, May 19, to widen the reach of its latest Gemini tools across developer and consumer products. The company said Gemini 3.5 Flash is rolling out through the Gemini API in Google AI Studio, while Gemini Omni Flash is beginning to reach paid Gemini subscribers and Google’s Flow filmmaking tool. Google also used the event to update adoption figures for SynthID, its watermarking system for AI-generated media. The announcements tie together three parts of Google’s AI push: models for developers, media-generation tools for consumers and creators, and provenance tools meant to identify Google-made AI content. ### Where is Gemini 3.5 Flash actually showing up first? Google AI Studio said on May 19 that Gemini 3.5 Flash was rolling out through the Gemini API in AI Studio. The company framed the model as aimed at “agentic” work and complex coding tasks, a positioning that matches Google’s broader push at I/O toward tools that can handle multi-step workflows. Google’s own Gemini 3 Flash materials describe the model family as built for “frontier intelligence at speed” and say it is intended for agentic workflows, multimodal inputs and low-latency reasoning. The DeepMind product page says the model can manage large numbers of function calls and reason across text, audio, images, code and video. ### Why is Google talking so much about coding and agents? (blog.google) Google’s product pages for Gemini 3 Flash put coding near the center of the pitch. The company says the model can generate richer visualizations, support rapid iteration and handle multi-step tool use, including cases with large numbers of tools and inputs. Google’s November 2025 launch post for Gemini 3 said Gemini 2 had laid the groundwork for agentic capabilities and that Gemini 3 would extend that work across the Gemini app, AI Studio and Vertex AI. (deepmind.google) Tuesday’s rollout of a newer Flash variant in AI Studio fits that pattern by putting a faster model into the developer surface where Google already distributes its APIs and SDKs. ### What is Gemini Omni Flash, and who gets it now? Google said Gemini Omni Flash is also starting to roll out on May 19 to Google AI Plus, Pro and Ultra subscribers. According to Google’s social posts, the model is appearing in the Gemini app and in Flow by Google, the company’s AI filmmaking tool. Omni features are also due to reach YouTube Shorts and the YouTube Create app later this week. (blog.google) Flow is Google’s filmmaking product built around its generative media stack, including Veo, Imagen and Gemini. YouTube has already been adding AI creation features to Shorts, including Veo-powered remix tools and AI-generated effects, so Omni’s arrival extends an existing rollout into creator products rather than opening a new category. ### What do the new SynthID numbers tell us? (blog.google) Google said on May 19 that SynthID has now watermarked more than 100 billion images and videos and 60,000 years of audio, and that its detector has been used 50 million times. Those figures are a sharp increase from May 2025, when Google said more than 10 billion pieces of content had been watermarked and the SynthID Detector was only starting to roll out to early testers. (blog.google) Pushmeet Kohli, Google DeepMind’s vice president for science and strategic initiatives, said last year that SynthID was meant to help identify AI-generated content across image, audio, video and text formats made with Google tools. The detector works by scanning uploaded media for Google’s watermark and highlighting portions most likely to contain it. ### What should developers and creators watch next? (blog.google) Google I/O 2026 runs through May 20 at Shoreline Amphitheatre in Mountain View, California, and online through Google’s event site. Google said Omni features are scheduled to reach YouTube Shorts and YouTube Create this week, while Gemini 3 is also being shipped across Search, the Gemini app, AI Studio and Vertex AI. (blog.google 1) (blog.google 2)