Google introduces Imagen 2, a new DALL-E 3 competitor for text-to-image generation

Google has unveiled an upgrade to Google Cloud's image-generation capabilities with the introduction of Imagen 2, the latest in text-to-image technology. This advanced system is now accessible to Vertex AI customers.

Imagen 2, developed using Google DeepMind technology, offers enhanced image quality and a range of features that cater to developers' specific needs. These capabilities include generating high-resolution, photorealistic images from natural language prompts, text rendering in multiple languages for accurate text overlays in images, and logo generation for creating company or product logos.

Additionally, Imagen 2 is equipped with a visual question and answering feature that generates captions from images and provides informative text responses to queries about image details. Beyond English, the technology is launching with support for six additional languages (Chinese, Hindi, Japanese, Korean, Portuguese, Spanish) in preview, with a host of other languages planned for release in early 2024.

To ensure that the generated images align with Google’s Responsible AI principles, Imagen 2 includes built-in safety precautions. Customers using Imagen 2 on Vertex AI can customize and deploy the technology with intuitive tooling, fully-managed infrastructure, and built-in privacy and safety features.