Imagen 4 is noted for its speed, being faster than Imagen 3, with plans for a variant that will be up to 10 times quicker. The model is now available in the Gemini app, Google’s Whisk and Vertex AI platforms, and across Google Workspace applications like Slides, Vids, and Docs. This release positions Imagen 4 as a competitive option among AI image generators, which include tools from ChatGPT and Midjourney.
Key takeaways:
- Google has unveiled Imagen 4, a new image-generating AI model, at Google I/O 2025, claiming it delivers higher-quality results than its predecessor, Imagen 3.
- Imagen 4 is capable of rendering fine details and can handle both photorealistic and abstract styles, creating images in various aspect ratios and up to 2K resolution.
- The model is faster than Imagen 3, with plans for a variant that will be up to 10x quicker, and is available in the Gemini app, Google’s Whisk and Vertex AI platforms, and across Google Workspace.
- Josh Woodward from Google’s Labs group highlighted improvements in text and topography generation, making Imagen 4 suitable for creating slides, invitations, and other media blending imagery and text.