Generative Media | I/O 2026 Keynote

TL;DR: The Google I/O 2026 Keynote introduced major updates to three generative media products—Pics (image creation and editing), Stitch (UI design), and Google Flow (multimodal creation)—including Gemini Omni, agents, custom tools, and music remixing features, emphasizing how technology helps turn creativity into reality faster. ## Google Pics: New Image Tool in Workspace Leveraging the powerful capabilities of Nano Banana, Google launched **Google Pics**, an image creation and editing tool integrated into Google Workspace. It supports almost any content from party flyers to infographics, offering fine-grained creative control. - **Smart Understanding**: Pics can recognize the content in a piece and the interactions between objects. Users can hover over an element to remove or resize it with one click, fitting it into the frame. - **Text and Translation**: After adding or editing text, it can be translated in just a few clicks. - **Safe Watermarking**: All outputs from Pics come with a SynthID watermark. - **Availability**: Rolling out starting this summer. ## Stitch: Build UI at the Speed of Thought **Stitch**, a design product used internally at Google, is now available globally with a new version. Over the past year, users worldwide have generated over 100 million UI screens via Stitch. Now it offers a new way to design: - **Real‑time Generation**: With just one prompt, Stitch generates UI designs in real time. - **Real‑time Collaboration**: Users can collaborate with Stitch by typing prompts or speaking. For example: "Make the title text larger, update the menu to highlight more pizza options" – the layout updates instantly. - **Export and Publish**: Connects with multiple tools to export designs as code or publish a website with one click. - **Availability**: Available to global users starting today. ## Major Updates to Google Flow Since its launch at last year’s I/O, millions have used Google Flow. This year brings three big updates: Gemini Omni, new agents, custom tools, plus music remixing. ### Gemini Omni: Keep the Original Performance, Change Everything With a simple prompt and a style reference, Gemini Omni can change the environment, add visual effects, and other elements while fully preserving the original performance (e.g., walking posture, rhythm). Users can even add new characters while keeping everything else in the scene unchanged. ### New Agents: Execute Multiple Actions Simultaneously Previously Flow could only handle one prompt at a time; now agents can manage multiple tasks concurrently. Example: given an image, the agent analyzes the scene, determines the best camera angles, and turns one image into 16 unique videos. It can also handle large‑scale edits, such as transforming every scene from early morning to late night – the desert sky goes dark, headlights turn on and illuminate the dust. ### Flow Tools: Code Creative Tools Freely Users can code any creative tool they imagine inside Flow – like video effects, hand‑drawn animations, or text overlays – and customize it to their personal workflow. Starting today, users can build, share, and remix tools. ### Flow Music: Turn Raw Recordings into Original Songs Users can record a piano melody stuck in their head into Flow Music, prompt it to go in an R&B direction, and add a female vocal. In an example, a raw recording processed by Flow produced a demo to guide the band. It’s not the final track, but it helps the band decide on subsequent recording directions. **All new Flow and Flow Music features are available today.** ## Technology as a Canvas for Creativity The real breakthrough isn’t the technology itself, but what people create with it. From musicians to small businesses, from creative coders to artists, Google’s products help shorten the distance from a spark of inspiration to making it real. “You are in an era where humans must be the most creative.” Source: https://www.youtube.com/watch?v=FLynjUYg79I