2025-06-26 - https://developers.googleblog.com/en/imagen-4-now-available-in-the-gemini-api-and-google-ai-studio/

Google has introduced Imagen 4, its most advanced text-to-image generation model, now available for paid preview via the Gemini API and limited free testing on Google AI Studio. Imagen 4 offers substantial improvements in text rendering quality over previous models. The Imagen 4 family includes two versions: Imagen 4 for most use cases ($0.04 per image) and Imagen 4 Ultra ($0.06 per image), which delivers more precise alignment with text prompts. The blog highlights various creative examples generated by Imagen 4 Ultra. Additionally, all images include a non-visible SynthID watermark for trust and transparency. Developers can start building with Imagen by accessing official documentation and cookbooks, with general availability expected in the coming weeks.