Google Rolls Out Paid Preview for Imagen 4 and Imagen 4 Ultra in Gemini API and AI studio

June 24, 2025

Details

  • Google AI Developers launched a paid preview for its advanced text-to-image tool, Imagen 4, on June 24, 2025.
  • Developers can access the service via the Gemini API or the Google AI Studio, catering to both code and no-code users.
  • The launch features two models: Imagen 4 for standard image synthesis, and Imagen 4 Ultra, which provides outputs that more closely match user prompts.
  • According to Google, the Ultra version achieves improved resolution, enhanced color fidelity, and fewer artefacts compared to the Gemini-embedded Imagen 2.
  • Access is metered separately from Gemini token usage and requires activating generative images in Google Cloud.
  • The preview includes safety filters and watermarking, with plans for wider rollout later in 2025.

Impact

This paid launch puts Google in direct competition with OpenAI's DALL·E 4 and Adobe Firefly for the enterprise image API market. By charging during the preview, Google not only recoups GPU expenses but may also impact pricing industry-wide. Its integration with Google Cloud positions the service to meet privacy regulations such as the EU AI Act, while developer feedback will inform future advancements in generative AI for video and 3-D content.

Read more