OpenAI's DALL-E 2 Now Open to All: A Game-Changer in AI Art
Written on
The Availability of DALL-E 2
As of September 28, 2022, OpenAI has made its highly anticipated DALL-E 2 accessible to the public. The research organization announced in a blog entry that the previous waitlist is now removed, enabling users to start creating images immediately.
> “Starting today, we are removing the waitlist for the DALL·E beta so users can sign up and start using it right away. More than 1.5 million users are now actively creating over 2 million images daily with DALL·E—ranging from artists and creative directors to authors and architects—with over 100,000 users sharing their creations and feedback in our Discord community.” — OpenAI
To explore this AI tool, you can sign up here.
What is DALL-E 2?
DALL-E 2 is an AI-driven image generator built on advanced machine learning techniques, developed by OpenAI. This innovative tool allows users to create original artwork by simply entering text prompts.
In April, OpenAI introduced the beta version, quickly gaining recognition as one of the leading AI systems capable of bringing any imaginative concept to life.
For instance, using the text prompt: 3D render of a cute tropical fish in an aquarium on a dark blue background, digital art, the AI produces four high-resolution images (1024x1024) in under 10 seconds—a remarkable feat.
Pricing Structure
The pricing for DALL-E 2 has remained unchanged since the introduction of a credit-based system in July. Users can purchase credits at a rate of $15 for 115 generations, with a maximum purchase limit of $1,500.
Let's break down the costs:
115 credits x 4 images per prompt = 460 images
$15 / 460 images = approximately $0.032 per image
Comparing DALL-E 2 with Competitors
When evaluating DALL-E 2's pricing against competitors, we find the following:
- MidJourney: Basic membership is $10 for 200 generations per month; Standard membership is $30 for unlimited generations.
- Stable Diffusion: Free if run locally on your PC; DreamStudio membership is $10 for about 1000 generations.
OpenAI aims to uphold a reputation as a controlled and brand-friendly alternative to other more permissive tools like Stability AI. The latter has faced criticism for its open-source Stable Diffusion, which has been used to generate inappropriate content.
OpenAI has stated: “We have made our filters more robust at rejecting attempts to generate sexual, violent, and other content that violates our content policy, and building new detection and response techniques to stop misuse.”
Current limitations include:
- Generation of images involving famous personalities
- Explicit content, such as nudity or violence
- Spelling errors—while DALL-E 2 excels in artistry, it struggles with spelling words.
For detailed content policies, click here.
Future Prospects
While this development is promising, it may be perceived as too little, too late for OpenAI. During DALL-E 2's restricted access, competitors like MidJourney and Stable Diffusion advanced significantly, producing results of comparable or superior quality.
Moreover, MidJourney has actively experimented with various models, including a new integration with Stable Diffusion, enhancing coherence in results, improving facial features, and more, while OpenAI has been less visible in the community.
Despite the existing limitations and the late introduction to the public, DALL-E 2 remains a powerful tool capable of generating exceptional images. I encourage everyone to give it a try and see what amazing creations you can make.
Chapter 2: Video Insights
Explore a live reaction to the launch of DALL-E 2 by OpenAI and see firsthand how it works.
Check out this tutorial on how to get started with DALL-E 2, now available to everyone!