We're thrilled to bring Imagen 4, our best text-to-image model yet, to paid preview in the Gemini API and for limited free testing in Google AI Studio. Imagen 4 offers significantly improved text rendering over our prior image models and pushes the boundaries of text-to-image generation quality.
We’re introducing two models within the Imagen 4 family, built to serve a variety of creative needs:
This is our flagship text-to-image model designed to handle a wide range of image generation tasks with significant improvements in quality, particularly for text generation, over Imagen 3. Imagen 4 is priced at $0.04 per output image.
When you need your images to precisely follow instructions, Imagen 4 Ultra is the model for you. It's designed to produce outputs that are more highly aligned with your text prompts, achieving strong results compared to other leading image generation models. Imagen 4 Ultra is priced at $0.06 per output image.
We will introduce additional billing tiers in the coming weeks. In the meantime, you can request higher rate limits for Imagen 4 and 4 Ultra.
To give you a glimpse of Imagen 4's capabilities, here are some examples of what you can create. Created using Imagen 4 Ultra, the prompts below showcase the model's versatility across various styles and content.
Prompt: A 3-panel cosmic epic comic. Panel 1: Tiny 'Stardust' in nebula; radar shows anomaly (text 'ANOMALY DETECTED'), hull text 'stardust'. Pilot whispers. Panel 2: Bioluminescent leviathan emerges; console red text 'WARNING!. Panel 3: Leviathan chases ship through asteroids; console re text 'SHIELD CRITICAL!', screen text 'EVADE!'. Pilot screams, SFX 'CRUNCH!', 'ROOOOAAARR!'.
Prompt: Front of a vintage travel postcard for Kyoto: iconic pagoda under cherry blossoms, snow-capped mountains in distance, clear blue sky, vibrant colors.
Prompt: Photograph of an adventurous couple hiking on a mountain peak at sunrise, arms raised in triumph, epic panoramic view of valleys below, dramatic light.
Prompt: Avant-garde fashion editorial shot: a model in a voluminous, architectural gown standing on a shimmering, alien landscape under a binary sunset, surreal colors, high-concept, cinematic.
To help maintain trust and transparency, all images generated by Imagen 4 models will continue to include a non-visible digital SynthID watermark. To get started, check out the official documentation and cookbooks:
We can't wait to see what you create with Imagen 4 through the Gemini API and Google AI Studio, and we look forward to making these models generally available in the coming weeks.
Using KerasHub for easy end-to-end machine learning workflows with Hugging Face
Gemini 2.5 for robotics and embodied intelligence
Google Cloud donates A2A to Linux Foundation
Supercharge your notebooks: The new AI-first Google Colab is now available to everyone