Google is rolling out major upgrades to its AI video generation ecosystem. The company unveiled Veo 3.1, featuring richer audio generation, improved narrative control, and enhanced realism that captures true-to-life textures. The new model builds on Veo 3 with stronger prompt adherence and better audiovisual quality when converting images to video.

The updates extend to Flow, Google's AI filmmaking tool that has generated over 275 million videos in its first five months. For the first time, audio is available across all Flow features including "Ingredients to Video," "Frames to Video," and "Extend." Users can now control characters, objects, and style using multiple reference images, generate seamless transitions between starting and ending frames, and create longer clips lasting a minute or more.

New editing tools are also launching within Flow. The "Insert" feature lets users add elements like realistic details or fantastical creatures, with Flow handling complex details like shadows and lighting. A removal tool is coming soon that can seamlessly erase unwanted objects and reconstruct the background. The Veo 3.1 model is available via the Gemini API for developers and Vertex AI for enterprise customers, alongside the Gemini app.