Neura Market
DALL·E 3 logo

DALL·E 3

Free

Announcement of DALL·E 3 image generator. OpenAI blog, September 20, 2023.

FreeFree tier
Inputs: textOutputs: image
Type
Open Source
Company
OpenAI
Visit Website GitHub (40 stars)

About DALL·E 3

DALL·E 3 is a text-to-image generation model developed by OpenAI, built natively on ChatGPT to allow users to generate images that precisely adhere to their text prompts. It represents a significant improvement over DALL·E 2, capable of interpreting anything from a simple sentence to a detailed paragraph. Users can leverage ChatGPT as a brainstorming partner to refine prompts, automatically generate tailored descriptions, and make iterative tweaks to images. DALL·E 3 includes safety mitigations such as declining requests involving public figures, improving protections against harmful biases, and offering creators the ability to opt their images out of future training. The model is available to all ChatGPT users (with varying limits based on subscription plan) and to developers via OpenAI's API.

Key Features

Built natively on ChatGPT for seamless prompt engineering and refinement
Generates images that precisely adhere to text, from simple sentences to detailed paragraphs
ChatGPT automatically generates tailored, detailed prompts based on user ideas
Easily tweak and refine images with few words via ChatGPT
Safety mitigations: declines requests for public figure images and reduces harmful biases
Provenance classifier for identifying AI-generated images (in development)
Creative control: declines requests in the style of living artists, opt-out for training data
Available to all ChatGPT users (free and paid plans) and via API

Pros & Cons

Pros
  • Notable improvements in prompt adherence over DALL·E 2
  • Seamless integration with ChatGPT enables natural language interaction and iterative refinement
  • Safety measures proactively decline problematic requests (public figures, artist styles)
  • Creators can opt out of training data, providing attribution control
  • Available on multiple ChatGPT tiers, making it widely accessible
Cons
  • Image generation is limited and slower on the free ChatGPT plan
  • Full access and higher usage require a paid ChatGPT subscription (Plus or Pro)
  • Safety mitigations may still miss nuanced harmful content despite improvements
  • Provenance classifier for identifying AI-generated images is still experimental

Best For

Brainstorming and visualizing creative concepts with ChatGPT as a collaboratorGenerating custom images for presentations, marketing, or personal projectsRapid prototyping of visual ideas without requiring prompt engineering skillsCreating illustrations, artwork, or photorealistic images for storytelling and content creation

FAQ

Is DALL·E 3 free to use?
DALL·E 3 is available to all ChatGPT users, including those on the free plan. However, free plan users have limited and slower image generation. More extensive use requires a ChatGPT Plus or Pro subscription.
How does DALL·E 3 differ from DALL·E 2?
DALL·E 3 represents a leap forward in text-to-image generation, with significantly improved ability to adhere exactly to the provided text, even with the same prompts. It is also built natively on ChatGPT for easier prompt refinement.
Can I create images of public figures with DALL·E 3?
DALL·E 3 has mitigations that decline requests asking for an image of a public figure by name, to reduce potential misuse.
How can I opt my images out of training?
Creators can opt their images out from training of future image generation models through OpenAI's opt-out mechanism, available on the DALL·E 3 page.