AI text-to-image
New accounts receive 10 welcome credits
Anyone with an image link can view reference uploads and generated images.
AI text-to-image: describe a specific scene instead of stacking adjectives
Text-to-image creates a new image from a written description. Describe the subject and setting first, then the lighting, viewpoint, and layout. Start with inspiration for product photography, café content, or landscape concepts.
How do text-to-image and image editing differ?
Text-to-image starts from a description and needs no original image. Image editing uses an existing image as a reference for changes to the background or atmosphere. For consistent real products, use a clear reference and check each detail afterward.
Can I generate images now?
Sign in to create with GPT Image 2 or GPT Image 2.5. 1K images cost 2 credits each. 2K and 4K are not available yet. New accounts receive 10 credits. View and download your creations in History.
Will my reference image be uploaded?
Your selected image is previewed only on this page. Clicking Generate submits it to the image service for editing. The original reference image is not saved to your library.