At Google I/O, Google Cloud has raised the bar for its generative AI capabilities with the introduction of the latest media models, Imagen 4, Veo 3, and Lyria 2 which are now available on Vertex AI. These powerful updates empower creators to generate the highest quality images, videos, and audio content using AI.
Imagen 4: Next Level of Image Generation
Google has introduced Imagen 4, its highest-quality text-to-image generation model, now available in public preview on Vertex AI.Key improvements:
- Outstanding text rendering with improved alignment to prompt instructions
- Enhanced image quality
- Supporting global creators with multiple languages prompts
Veo 3: Higher-quality video generation with audio and speech
Generate videos with the latest video generation model from Google Deep Mind. Key improvements:
- Improved quality of video output
- Natural-sounding speech synthesis
- Richer audio quality, such as music and sound effects
- All of this is generated simply from a text prompt
Lyria 2: Dynamic and Customizable Music Generation
Lyria, Google’s text-to-music model was announced at Google Cloud Next 2025. Google Cloud has updated the latest version with higher quality audio and control.Key improvements:
- Higher quality audio output from simple text prompts
- More control over musical elements like instruments, BPM, and arrangement
- Greater creative flexibility for composers
📢Get GCP Onboarding & AI Foundation Setup for free with GoPomelo!
Submit form below and kickstart your AI innovation with GoPomelo’s complimentary onboarding and foundational setup services for organizations looking to explore generative AI on Google Cloud Platform (GCP).
What can GoPomelo Offer?- GCP Local Billing Setup - Streamline GCP usage with local currency billing
- AI-Ready Foundation Setup - We set up the components to help you build securely and scale effortlessly including:
-
- Google Identity Integration
- GCP Project Setup
- IAM (Identity & Access Management) Configuration
-
Source: Google Cloud Blog