Google is in a filmmaking frame of mind. The search giant introduced Veo 3, the latest version of its generative video model, loading it with cinematic capabilities including a new AI storytelling tool called Flow. At the Google I/O conference the company also debuted an upgraded image generator, Imagen 4, and announced expanded access to the AI music tool Lyria 2. Veo 3 can generate videos with audio — a Google first, adding things like background traffic noises, birds singing, “even dialogue between characters.” It offers improved consistency of characters, scenes and objects, while gaining camera controls, outpainting and object add/remove.

“We’ve partnered closely with the creative industries — filmmakers, musicians, artists, YouTube creators — to help shape these models and products responsibly and to give creators new tools to realize the possibilities of AI in their art,” Google explains in a blog post.

The company simultaneously announced a partnership with director Darren Aronofsky, whose Primordial Soup is producing three shorts using Google’s AI tools.

“Veo 3 and Imagen 4 generative AI models crank the realism dial to 11,” writes PetaPixel, noting “Veo 3 excels at generating text and photorealistic scenes and is better able to replicate real-world physics than prior Veo models,” and can also lip sync.

Veo 3 is available in the U.S. to subscribers of the new Google AI Ultra service, which costs $249.99 per month (with an introductory offer of 50 percent off for three months). Ultra subscribers can access Veo 3 in the Gemini app and in Flow. As with all the new tools, enterprise users can also access it on Vertex AI.

“For the first time, we’re emerging from the silent era of video generation. [You can give Veo 3] a prompt describing characters and an environment and suggest dialogue with a description of how you want it to sound,” said Google DeepMind CEO Demis Hassabis, per TechCrunch.

SiliconANGLE reports Flow “integrates Google’s AI capabilities — Veo for video, Imagen for visual assets, and Gemini for natural language understanding — into a single collaborative interface.”

With Flow, “users can create scenes using natural prompts, manage ingredients like cast, settings and objects and edit storylines fluidly with tools such as SceneBuilder and camera controls,” SiliconANGLE explains, noting “a built-in feature called Flow TV showcases clips created by others to offer inspiration.”

Google in April began providing access to its Lyria 2 music generator and now makes it available to creators in YouTube Shorts. Lyria RealTime allows users to “interactively create, control, and perform generative music in real time.” Lyria is available via API and in AI Studio.

Available in the Gemini, Whisk and across Slides, Vids and Docs in Workspace, Imagen 4 has improved spelling and typography and offers finer detail at images at up to 2K, with a 10x faster version coming soon, Google says.

Related:
Creating in Flow: How to Use Google’s New AI Filmmaking Tool (Video), Google, 5/21/25
Google’s Veo 3 Is Already Deepfaking All of YouTube’s Most Smooth-Brained Content, Gizmodo, 5/22/25