How do I create an image and find its result?

Use the Image workbench to generate an image from a description or work with reference images supported by the selected model.

1. Open Image and enter your description. Choose an available model and its settings.
2. If using reference images, upload them and wait until they are ready. Check the reference limits for that model.
3. Check the current cost and generate. Open the completed result to inspect it and use its download action.
4. To return later, open the Image project records and select the project and generation round.

If uploading, generating or downloading fails, tell Support the model, project, time and failing action. Model settings and supported reference options vary; use the current page’s options. Do not create another generation just to locate an existing result.

Images generated on Free carry a MyVocal watermark and count toward a daily image allowance. Paid generations do not add that watermark. The number of images left today is separate from the character quote for this generation. For a batch, check each image: reserved, charged and released characters are different, and releasing characters is not a cash refund.

The current Image model catalog includes OpenAI’s GPT Image 2, GPT Image 2.5 Flare and GPT Image 2.5 Sunburst, Google’s Nano Banana, Nano Banana Pro, Nano Banana 2 and Nano Banana 2 Lite, and ByteDance’s Seedream 5 Lite and Seedream 5 Pro. Access to a model and its settings depends on your plan and the options currently offered in the Image workspace. The quote shown for the current operation determines the generation cost.