Google Unveils Next Gen AI Creative Tools
On May 20, 2025, Google announced significant breakthroughs with its newest generative media models. These models are designed to create breathtaking images, videos, and music, empowering artists to bring their creative visions to life and providing powerful tools for everyone to express themselves.
The new lineup includes Veo 3 and Imagen 4, cutting-edge video and image generation models, respectively. Google is also expanding access to Lyria 2 for musicians and inviting visual storytellers to try Flow, a new AI filmmaking tool. Powered by Google DeepMind’s most advanced models, Flow enables users to craft cinematic films with sophisticated control over characters, scenes, and styles.
These developments are the result of close partnerships with the creative industries, including filmmakers, musicians, artists, and YouTube creators. This collaboration helps shape these models and products responsibly, ensuring creators have new tools to explore the possibilities of AI in their art. You can learn more about these collaborations with filmmakers and how they shape these models.
Veo 3: Revolutionizing Video Generation with Integrated Audio
Veo 3, Google's new state-of-the-art video generation model, not only improves upon the quality of Veo 2 but, for the first time, can also generate videos with audio. This includes environmental sounds like traffic noises in a city scene, birds singing in a park, or even dialogue between characters.
Veo 3 excels across various aspects, from text and image prompting to understanding real-world physics and delivering accurate lip-syncing. It's adept at interpreting narrative prompts, transforming short stories into vivid video clips. Veo 3 is available today for Ultra subscribers in the United States in the Gemini app and in Flow. It is also accessible for enterprise users on Vertex AI.
Veo 2 Enhancements: Empowering Filmmakers with Advanced Control
As Veo 3 advances, Google has also enhanced its popular Veo 2 model with new capabilities, developed in collaboration with creators and filmmakers. Several of these new features are launching today:
- State-of-the-art reference-powered video: This capability allows users to provide Veo with images of characters, scenes, objects, and even styles for improved creative control and consistency.
- Camera controls: Define precise camera movements, including rotations, dollies, and zooms, to achieve the perfect shot.
- Outpainting: Broaden your video frame, for example, by turning portrait footage into landscape. This feature makes it easier to fit any screen size by intelligently adding to the scene.
- Object add and remove: Add or erase objects from videos. Veo understands scale, interactions, and shadows, using this understanding to create natural, realistic-looking scenes.
Reference-powered video and camera controls are now available in Flow. Google plans to bring all these new Veo 2 capabilities to the Vertex AI API in the coming weeks, and to more products over the next few months.
Veo 2 Examples:
Video Example: Veo 2 camera controls Watch Video
Video Example: A woman walking in a hallway made with Veo2 Watch Video
Video Example: A knit scene made by Veo (Original) Watch Original Video
Video Example: A knit scene made by Veo (Outpainted) Watch Outpainted Video
Video Example: Astronaut scene with Veo (Original) Watch Original Video
Video Example: Astronaut scene with Veo (Spaceship Removed) Watch Modified Video
Introducing Flow: Your AI Co-pilot for Cinematic Storytelling
Built with and for creatives, Flow is an AI filmmaking tool designed to seamlessly create cinematic clips, scenes, and stories. It brings together Google DeepMind’s most advanced models: Veo, Imagen, and Gemini. Users can describe shots to Flow using natural language, manage story ingredients like cast, locations, objects, and styles in a single convenient place, and use Flow to weave narratives into beautiful scenes.
Flow is available today for Google AI Pro and Ultra plan subscribers in the U.S., with more countries coming soon.
Watch: Flow product sizzle video
Imagen 4: Pushing the Boundaries of Image Quality and Typography
Google's latest image model, Imagen 4, combines speed with precision to create stunning images. It boasts remarkable clarity in fine details such as intricate fabrics, water droplets, and animal fur, and excels in both photorealistic and abstract styles. Imagen 4 can generate images in a range of aspect ratios and up to 2K resolution, making it even better for printing or presentations. Furthermore, it is significantly improved at spelling and typography, simplifying the creation of custom greeting cards, posters, and even comics.
Imagen 4 Examples:
Imagen 4 is available today in the Gemini app, Whisk, Vertex AI, and across Slides, Vids, Docs, and more in Workspace.
Soon, Google will also launch a fast variant of Imagen 4 that is up to 10x faster than Imagen 3, allowing for even quicker idea exploration.
Lyria 2: Expanding Musical Creativity with AI Tools
In April, Google expanded access to Music AI Sandbox, powered by Lyria 2. Music AI Sandbox offers musicians, producers, and songwriters a set of experimental tools designed to spark new creative possibilities and help artists explore unique musical ideas. Feedback from the music industry helps ensure these tools empower creators and invite them to realize AI's potential in their art.
Lyria 2 brings powerful composition and endless exploration capabilities. It is now available for creators through YouTube Shorts and for enterprises in Vertex AI. Additionally, Lyria RealTime, an interactive music generation model powering MusicFX DJ, is available via an API and in AI Studio. Lyria RealTime allows anyone to interactively create, control, and perform generative music in real time.
Commitment to Responsible AI Creation and Collaboration
Since its launch in 2023, SynthID has watermarked over 10 billion images, videos, audio files, and texts. This helps identify them as AI-generated and reduces the chances of misinformation and misattribution. Outputs generated by Veo 3, Imagen 4, and Lyria 2 will continue to feature SynthID watermarks.
Today, Google is launching SynthID Detector, a verification portal to help people identify AI-generated content. Users can upload a piece of content, and the SynthID Detector will identify if the entire file or just a part of it contains SynthID.
With all its generative AI models, Google aims to unleash human creativity and enable artists and creators to bring their ideas to life faster and more easily than ever before.
