DALL·E 2

Model ● Covered in 4 stories + Follow

DALL·E 2 is a text-to-image generation model developed by OpenAI that uses diffusion processes guided by text conditioning through CLIP embeddings to create images from text descriptions. Recent work has focused on improving the model's safety and fairness, including implementing demographic diversity techniques to reduce bias in generated images of people and deploying content policy guardrails during training. The model has been shaped through feedback from over 3,000 artists across 118 countries during its research preview phase before broader availability.

Updated 8 August 2026

Specifications

No specifications recorded yet.

Latest developments

Text-to-Image: Diffusion, Text Conditioning, Guidance, Latent Space Eugene Yan · 3 years ago · 20

Reducing bias and improving safety in DALL·E 2 OpenAI · 4 years ago · 18

DALL·E 2: Extending creativity OpenAI · 4 years ago · 50

DALL·E 2 pre-training mitigations OpenAI · 4 years ago · 25

Timeline

Month Quarter Year

2022

27 Nov Text-to-Image: Diffusion, Text Conditioning, Guidance, Latent Space
18 Jul Reducing bias and improving safety in DALL·E 2
14 Jul DALL·E 2: Extending creativity
28 Jun DALL·E 2 pre-training mitigations

Relationships

Products & technology

OpenAI develops this model · 3 sources

Every AI story that matters,
in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.