OpenAI introduces DALL-E 3, its newest text-to-image generation tool
OpenAI has announced the latest version of its text-to-image generation model, DALL-E 3, featuring integration with ChatGPT. This new combination allows users to craft detailed prompts and bring their ideas to life through accurate image generation. OpenAI claims that DALL-E 3 represents a significant leap forward in generating images that adhere closely to the provided text, addressing the issue of modern text-to-image systems ignoring certain words or descriptions.
Availability and access for different users
DALL-E 3 is currently in research preview and will be made available to ChatGPT Plus and Enterprise users in October. API and Labs customers will gain access later in the year. However, OpenAI has not announced when the latest text-to-image generating tool will be available for free customers. Similar to DALL-E 2, users do not need OpenAI's permission to reprint, sell, or merchandise images created using DALL-E 3.
Improvements and safeguards in DALL-E 3
OpenAI has implemented improved safeguards in DALL-E 3, addressing concerns such as generating images of public figures, propagating misinformation, and creating violent, adult, or hateful content. Additionally, the model is designed to decline requests for generating images in the style of living artists. Creators can also opt their images out of the training process for future image generation models.
Promised enhancements over DALL-E 2
The Sam Altman-led company promises that DALL-E 3 will deliver significant improvements over its predecessor, even when using the same text prompt. OpenAI said in a statement about its latest generative AI-based offering, "Modern text-to-image systems have a tendency to ignore words or descriptions, forcing users to learn prompt engineering. DALL·E 3 represents a leap forward in our ability to generate images that exactly adhere to the text you provide."