DALL-E 3

"Improving Image Generation with Better Captions" — the third generation of DALL-E, integrated natively into ChatGPT. Key innovation is a captioning model that generates detailed, accurate descriptions of training images, dramatically improving text-image alignment.

DALL-E 3 generates images that closely follow complex, detailed prompts without requiring prompt engineering. Users interact through natural conversation in ChatGPT rather than crafting specialized prompts. By Betker et al. Proprietary.