OpenAI launches DALL-E API so developers can integrate its AI illustrations into their applications

Enlarge / Variations of "a photo of 'a happy puppy corgi with fancy sunglasses sitting and facing forward, studio light, longshot' generated by OpenAI's DALL-E. Open AI

On Thursday, OpenAI announced the introduction of an API for its DALL-E image synthesis model that will allow developers to easily integrate its AI image generation technology into their applications. DALL-E, currently available as a standalone commercial service, allows users to generate new 1024x1024 images from textual descriptions called "prompts". Generated images can vary in style, from photorealistic to abstract, and many styles in between.

OpenAI's new API allows software developers to introduce image prompting and other parameters into OpenAI's DALL-E engine over the Internet. OpenAI then generates the image on its servers and returns the image to display it in the requesting application. (An API is a programming interface that allows different programs to communicate with each other.)

Using the API requires registration with OpenAI and a private API key allowing access to the DALL-E generator. Additionally, OpenAI charges a fee for each generated image based on the resolution of the image. 1024x1024 images cost $0.02 per image, 512x512 images cost $0.018 per image, and 256x256 images cost $0.016 per image.

Enlarge / A DALL-E example of "an astronaut on horseback". Open AI

Microsoft is already using the DALL-E API in closed-access trials of its new Microsoft Designer application, and Shutterstock recently announced the integration of DALL-E into its website. Additionally, OpenAI says design applications called CALA and Mixtiles also use the DALL-E API.

Since its launch in April, DALL-E and other computer-generated imagery models, such as Stable Diffusion, have been subject to ethical controversy due to built-in biases, the ability to potentially generate images harmful and that these AI generative models have been trained using human-made artwork without the consent of the individual artist. However, OpenAI implements filters that restrict the generation of nudity, "hate symbols", and violence in an effort to keep the service acceptable to a mass audience.

It should be noted that Stable Diffusion has been available as open source since August, and developers have already integrated this technology into their applications without the individual image generation costs, filters, or content policy restrictions that come with it. the OpenAI API DALL-E i...