# What is Sora? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Generative AI  
> **Canonical Citation:** [https://spidits.com/ai-glossary/sora](https://spidits.com/ai-glossary/sora)

## Definition
Sora is an advanced text-to-video diffusion model developed by OpenAI, capable of generating high-fidelity, photorealistic video clips up to 60 seconds long from written text prompts.

## Architectural Mechanics & Deep Dive
Sora is an advanced text-to-video generative AI model. Built using a diffusion transformer (DiT) architecture, Sora processes visual data as space-time patches to generate highly realistic, complex scenes with multiple characters, specific motion types, and accurate background details, representing a major leap in physical world simulation.

## Industry Applications & Real-World Use Cases
Generative AI content creation, text-to-video synthesis, and creative prototyping.

## Frequently Asked Questions

### How does Sora generate videos?
Sora uses a transformer architecture over spacetime patches of video data, learning to gradually denoise random pixels into structured video frames.

### What is a spacetime patch in Sora?
Spacetime patches are analogous to tokens in language models; they represent small 3D cubes of space and time data extracted from video frames.

## Related Terminology
- [openai](https://spidits.com/ai-glossary/openai)
- [diffusion-model](https://spidits.com/ai-glossary/diffusion-model)
- [generative-ai](https://spidits.com/ai-glossary/generative-ai)

---
*From the SPIDITS AI Knowledge Base (336+ terms indexed). Explore full technical definitions: [https://spidits.com/ai-glossary](https://spidits.com/ai-glossary)*