Following the huge success of DALL-E as a commercial service for image generation from text prompts, OpenAI has released an API that allows developers to easily integrate its into their apps. 

As we've previously reported, DALL-E, which made its debut in 2021, produces art via text. Specifically it uses a combination of OpenAI's GPT-3 transformer architecture with CLIP, a neural network which learns visual concepts from natural language supervision. DALL-E was trained using 250 million text-image pairs taken from the internet, mostly from Wikipedia and learned to generate new images based on text inputs.

According to OpenAI in the announcement of the new API:

More than 3 million people are already using DALL·E to extend their creativity and speed up their workflows, generating over 4 million images a day. Developers can start building with this same technology in a matter of minutes.

The Images API is remarkably simple - you provide an image prompt and other parameters into OpenAI's DALL-E engine over the Internet.

OpenAI then generates the image on its servers and returns the image for display in the requesting app. The prompt in the snippet above returns the top image in this example from the Image Generation documentation. The prompt for the lower one is more elaborate:

As well as creating images from scratch based on a text prompt, the API also lets you create edits of an existing image based on a new text prompt and this is one of the variations by the text:

"a photo of a happy corgi puppy with fancy sunglasses on sitting and facing forward, studio light, longshot" 

To use the API you'll first need to register with OpenAI in order to be given a private API Ket that allows access to the DALL-E generator. OpenAI levies a fee for each image generated depending on image resolution. Images that are 1024×1024 cost $0.02 an image, 512×512 images cost $0.018 per image, and 256×256 images cost $0.016 per image.

The Images API is in beta during which the API and models will evolve based on feedback. To ensure all users can prototype comfortably, the default rate limit is 10 images per minute, 25 per 5 minutes. The default rate limit will be increased as OpenAI learns more about usage and capacity requirements. 

Microsoft is among the early access partners who have already been working with DALL-E. Its plans include integrating DALL·E in Bing and Microsoft Edge with Image Creator, allowing users to create images if web results don’t return what they’re looking for and in a new graphic design app called Designer, which will helps users create professional quality social media posts, invitations, digital postcards, graphics, and more and you can apply to join the Designer wait list.

More Information

DALL-E API Now available in public beta

Image generation documentation

Related Articles

A Tale Of Two AIs - Ai-Da And DALL-E

The Unreasonable Effectiveness Of GPT-3

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.