Leverage our prompt builder to generate the visuals closest to your vision. Start by entering a description of the visuals you wish to see in the prompt field. Use the blueprint below as a guide for creating effective prompts:  
  
[Main Subject] [Action] [Surroundings] [Visual Aesthetics]   
  
The following are prompt suggestions to add to any query: high quality, high fidelity, detailed, intricate, sharp, proportional.    
  
Turn on negative prompt to specify what you don’t want to see in an image. For example, “clouds”, or “muted colors”.  
  
Visit our User Guide for more guidance and inspiration.

Focus on what you want to see in your generated visuals: subject, medium, environment, lighting, color, mood, composition. Use the “Negative prompt” toggle to define what you do not want to see in your generated image. 

Do you have an image from your Generation History, a licensed image from our creative library, or your own image* you want to use for inspiration? Upload it as a Reference Image so the model can follow the color or composition in its generation. 

*If you elect to upload an image, you will be prompted to confirm you own or hold a license to the image to ensure commercially safe generated images. Images uploaded as a Reference Image are not added to Getty Images’ creative library or used for AI model training. Supported files include JPG, JPEG, PNG, and WebP with a maximum file size limit of 30MB with a minimum dimension of 1024px.

We suggest thinking of your positive prompt as two parts: natural language first; then add keywords, separated by commas. Include keywords that will reinforce the concept described in your natural‑language phrase.   
  
Here’s an example:   
A white cat sleeps on a blue textured pillow. It is a wide‑angle shot, and you can see the entire room. Cat, fluffy, pillow, texture, detail, interior, wide angle  
  
For your negative prompt, natural language is not recommended. Instead, enter a list of keywords (separated by commas) that describe what you do not want to see in your generated image. Using the example above, consider writing the following negative prompt: dog, outdoors.   
  
Note: If you add “no” in front of a word in the negative prompt space, it will be read as a double negative and will instead include the phrase you meant to exclude.

If there is an image that you have in mind for color or composition, you can upload that as a Reference Image – from your Generation History, previously licensed from our creative library, or one that you own or have the license to use.

Yes, text prompt entry is available in 75 languages, empowering worldwide creativity. Supported languages include: Afrikaans, Albanian, Amharic, Arabic, Armenian, Azerbaijani, Bengali, Bosnian, Bulgarian, Catalan, Chinese (Simplified), Chinese (Traditional), Croatian, Czech, Danish, Dari, Dutch, English, Estonian, Farsi (Persian), Filipino, Tagalog, Finnish, French, French (Canada), Georgian, German, Greek, Gujarati, Haitian Creole, Hausa, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Irish, Italian, Japanese, Kannada, Kazakh, Korean, Latvian, Lithuanian, Macedonian, Malay, Malayalam, Maltese, Marathi, Mongolian, Norwegian (Bokmål), Pashto, Polish, Portuguese (Brazil), Portuguese (Portugal), Punjabi, Romanian, Russian, Serbian, Sinhala, Slovak, Slovenian, Somali, Spanish, Spanish (Mexico), Swahili, Swedish, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, Uzbek,   
Vietnamese, and Welsh.

Generative AI by Getty Images is powered by NVIDIA Edify, a custom machine learning model, trained exclusively using high‑quality content and proprietary data from Getty Images’ creative library. Customers then use text prompt‑to‑image generation to direct the model in creating imagery based on our content and data.

AI image generation is the process of creating never‑before‑seen images using artificial intelligence (AI) technology trained on a large data set. Getty Images' AI Generator, powered by NVIDIA Edify, was trained exclusively on our own creative image library and proprietary data, and allows customers to enter text prompts that the AI model then uses to generate four unique images that are commercially safe and ready to license.

Text prompt‑to‑image generation lets you create unique images in seconds by describing the image that you want to see. The AI Generator will then generate visuals relevant to those prompts that can be used across your creative process, from ideation to execution.

Once you have an active AI Generator agreement, open Getty Images’ AI Generator and type a description of what you want to generate into the text field provided. Your prompt can be up to 250 words long. The more details you provide, the better your results will be. 

Need help constructing a prompt? Try our innovative prompt builder, which guides you through the process by having you fill in specific fields. 

Want to generate an image that is similar in composition or color to one you have in mind? Use the Reference Image upload function to select an image from your Generation History, a previously licensed image from our creative library, or one that you own or have the license to use.

Once completed to your satisfaction, click "BUILD PROMPT," and the tool will transform the individual descriptions into a single prompt, ready to generate imagery.

AI image generation on Getty Images is not free. Each generation uses one generation credit from your purchased Generative AI agreement. This ensures we can continue to support creators and artists on an ongoing basis by compensating them for allowing their work to be part of our AI model.

No, AI art or image generators are not illegal. However, the context in which they are trained, and the data used to train them can determine if their output respects legally protected intellectual property rights. Questions to consider:  

• Was the AI model trained on data that did not have the proper permissions or artists' consent?  
• Do any of the AI‑generated images infringe on copyright or other intellectual property rights like trademark or rights of publicity/privacy? 

Some competitor AI image generators do, in fact, use data scraped from the internet. The visuals they generate run the risk of violating the intellectual property rights of content creators and brand IPs. This creates the risk of infringement liability for anyone using them.   
 
That's why we are vigilant in ensuring our AI model is only trained on Getty Images' creative library and proprietary data. We put in place safeguards to significantly decrease the risk of generating problematic visuals, so the images you generate are immediately safe and cleared for commercial use.  

Getty Images AI‑generated images include the same standard licensing agreement offered with our creative images. It allows for usage in a wide variety of projects including social media, online promotions, print marketing, and more. For additional rights, including broader sharing rights and uncapped legal indemnification, please visit our Overview & Pricing page for more information.

The Getty Images AI Generator console includes the ability to add a reference image to give you more precise control to influence the color or composition of your new generation. On the left side of the prompt bar, select "Add image," and choose "Add a reference image." From the module, you can add a reference image from your generation history, a licensed image from our creative library, or upload an image in the console. 

Note: If you elect to upload an image, you will be prompted to confirm you own or hold a license to the image to ensure commercially safe generated images. 

The Getty Images AI Generator allows you to upload your existing product image within the console and create new backgrounds that are tailored for your needs. Our AI technology removes the background automatically and allows you to adjust the product image size and position. Then, you can prompt and generate a new background scene while perfectly maintaining the uploaded object in the scene. 

To start, select “Add image” from the left side of the prompt bar and choose “Upload product image.” Supported files include JPG, JPEG, PNG, and WebP with a maximum file size limit of 30MB. You will be prompted to confirm you own or hold a license to the image to ensure commercially safe generated images. Images uploaded are not added to Getty Images’ creative library or used for AI model training.

We currently support JPG images across many aspect ratios and sizes, including Medium (1024x1024) and Large (4096x4096). Pixel density is at 72dpi.    

No, unlike some other AI services, we will not be adding visuals you generate or modify to our creative library for others to license and use.  

Generating images using the AI Generator grants you a license backed by our legal indemnification set forth in your applicable Generative AI Agreement. Please refer to our Content License Agreement and discuss any concerns you may have with your legal team. 

Yes! Content generated by the AI Generator grants you a licensed image with Getty Images’ automatic inclusion of legal indemnification.