Artificial intelligence · 2024

Sora

Minute-long, coherent video from a text prompt.

OpenAI previewed Sora on February 15, 2024: a diffusion transformer over spacetime patches generating up to a minute of high-fidelity video. It showed emergent 3D consistency and object permanence, and launched publicly in December 2024.

See it in the living tree Wikipedia

Arcs above join it to what it stood on, arcs below to what it led to. The bar places this stretch on the tree's whole timeline.

What it built on 3

DALL-E 2 2022
Denoising diffusion (DDPM) 2020
Vision transformer 2020

What it led to 1

Sora 2 2025

Watch

Introducing Sora — OpenAI’s text-to-video modelOpenAI · YouTube

Related measurements

The dashed line marks when it first worked.

Training compute of notable AI models Source: Epoch AI, Notable AI Models (CC BY 4.0)

ML hardware FLOP/s per dollar Source: Epoch AI, ML hardware (CC BY 4.0)