Turn a text prompt into an image, or add reference images on supported models. Sign in with Google; eligible new accounts get 5 free credits once after a security check. Check the credit cost before you generate.
Try these examples
From text to image
Use ImgGenAI to build a first draft or visual concept. Describe your subject, lighting, and composition. Choose a style preset with Z Image Turbo, or use AI Enhance for detail. Review the credit cost before each attempt.
01 | Start with a text prompt
Describe a ceramic cup or a mountain cabin. Tell the AI image maker about the setting and mood. Choose an aspect ratio and a 1K or 2K resolution depending on the model. Each request produces one image; check the Generate button for the exact credit cost.
02 | Add a reference image
Communicate a composition with a reference image on supported models. Upload a JPG, PNG, or WebP up to 10 MB per file. References are optional for Nano Banana Beta and GPT Image 2, but required for Qwen Image Edit Plus. Results vary, so review carefully.
03 | Review and refine
Look at the colors and details in your generated image, then refine your next prompt. Signed-in users can revisit their history in My Creations. Each new generation uses credits; check the cost and download the files you want to keep before the retention period ends.
AI Image
Compare the enabled models below, then choose one in the workspace. Available ratios, resolutions, reference options, and credit costs depend on the model.
Loading model catalog
Loading model catalog
Loading model catalog
Browse these examples for ideas about subjects and composition. Write your own description in the workspace above.
White ceramic candle on pale stone, soft daylight, clean premium commercial photography.
Fresh green serum bottle on stone counter, natural light, premium ecommerce campaign.
Fresh strawberries in a woven basket, warm daylight, cheerful food brand photography.
Soft natural-light portrait against a blurred green background.
Rainy Seoul alley at night with neon signage and wet street reflections.
Warm watercolor illustration of students gathered on a school field.
How to get started
Sign in with Google. Write what you want to see, including the subject, setting, and details about lighting or colors. You can use AI Enhance to elaborate your text.
Select a model, aspect ratio, and available resolution. Add a reference image on supported models and check the credit cost. Eligible new accounts get 5 free credits once after a security check.
Submit your prompt and review the single generated image. Download it or plan another prompt. You can revisit images in My Creations during the applicable retention period.
Explore realistic photo concepts before planning a shoot. Describe the materials, surface, lighting, and shadows, then review whether the generated details fit your idea.
Try a vertical cover concept or a wide event graphic. Choose a matching aspect ratio and leave room in your composition for text you plan to add later.
Compare color palettes and visual themes. Use AI image generation tools to discuss a visual direction with your team before committing to final assets.
Sketch out a story setting or a playful scene in words. Experiment with a style preset on Z Image Turbo and refine your description based on the output.
An AI image generator creates an image based on your text prompt. ImgGenAI lets you select a model, write a description, adjust settings, and review the result. You can use reference images on specific models, or try AI Enhance and style presets to develop a prompt.
You can try this free AI image generator with 5 credits: eligible new accounts get them once after Google sign-in and a security check. That can cover two separate Z Image Turbo generations at 2 credits each, or one Krea 2K image at 4 credits. Costs vary by model, resolution, and references. GPT Image 2 starts at 6 credits, so the trial balance does not cover it.
Enabled models include Z Image Turbo, Krea 2 Turbo, Nano Banana Beta, Wan 2.5 Image, Qwen Image Edit Plus, and GPT Image 2. Choose a model to see its available aspect ratios, resolutions, and exact credit cost.
Yes, on supported models. Upload JPG, PNG, or WebP files up to 10 MB each. Nano Banana Beta accepts up to 5 optional references; GPT Image 2 accepts up to 3. Qwen Image Edit Plus requires 1 to 3 reference images. Use only images you have permission to upload.
Available ratios include 1:1, 16:9, 9:16, 3:2, 2:3, 3:4, and 4:3, with 4:5 on some models. All enabled models offer 1K resolution. Krea 2 Turbo and Qwen Image Edit Plus also offer 2K. The choices shown depend on your selected model.
Your generation history is available in My Creations after sign-in. Image files may be scheduled for deletion after about 7 days for free access or about 90 days for paid access. Image retention is separate from credit validity. Download the images you want to keep.
Commercial use depends on the Terms of Service and applicable law. Review them before publishing, and check any rights related to your prompt, reference images, and output. ImgGenAI does not guarantee that a generated image is cleared for commercial use.