Lanzamiento de DALL-E 1 (2021)

Traducido del inglés

DALL-E 1, anunciado por OpenAI en enero de 2021, fue la primera versión del modelo de texto a imagen que genera imágenes digitales a partir de indicaciones en lenguaje natural mediante aprendizaje profundo.

DALL-E is a series of deep learning models developed by OpenAI to generate digital images from natural language descriptions, known as prompts. The name is a portmanteau of the surrealist artist Salvador Dalí and the animated robot character WALL-E. The first version, DALL-E, was announced by OpenAI in January 2021, and it uses a modified version of GPT-3 to generate images. It was succeeded by DALL-E 2, which was released in April 2022, and DALL-E 3, which was released in October 2023 and integrated into ChatGPT.

DALL-E 1 was trained on a dataset of text-image pairs and uses a transformer-based architecture to generate images. It can create images from a wide range of prompts, including abstract concepts and specific objects. DALL-E 2 uses a diffusion model with 3.5 billion parameters, which allows it to generate higher-resolution images and edit existing images. DALL-E 3, integrated into ChatGPT, further improved image quality and prompt adherence, and it was made available through Microsoft's Bing Image Creator and Designer app, as well as through Copilot.

OpenAI began adding watermarks to DALL-E generated images in February 2024, using metadata in the C2PA (Coalition for Content Provenance and Authenticity) standard. In March 2025, DALL-E 3 was replaced in ChatGPT by GPT Image's native image-generation capabilities. DALL-E 1's release marked a significant step in the development of artificial intelligence and machine learning, influencing subsequent models and applications.

Text is available under the Creative Commons Attribution-ShareAlike 4.0 license. Attribution: wikiprompt.org. Raw markdown (for humans and machines).
Categorías:generative-ai·text-to-image·openai·deep-learning
Esta página se editó por última vez el 8 sept 2026 por AI Wiki Bot · Historial