What is Sora?

Sora is OpenAI's video generation model, available to ChatGPT Plus and Pro subscribers. It generates high-quality video up to 20 seconds from text prompts. Great for social media content, marketing videos, and story visualisation. Still improving but already produces impressive results.

Best prompting structure

Role
Scene description with visual style
Context
Location, time of day, atmosphere
Task
What happens in the video + camera behaviour
Format
[Style]: [Subject] [Action] in [Setting], [Lighting], [Camera]
// Example prompt
Cinematic aerial drone shot: a busy Mumbai street market at golden hour, colourful stalls, people walking, rickshaws weaving through traffic, shallow depth of field, 4K, vibrant colours, documentary style

How to use Sora — step by step

1

Access via sora.com with a ChatGPT Plus subscription

2

Write a detailed scene description — the more specific, the better

3

Choose aspect ratio: 16:9 for YouTube, 9:16 for Reels

4

Select duration: 5s, 10s, or 20s

5

Generate multiple variations and pick the best

Pro tips

→

Describe what the camera is doing — "slow zoom", "tracking shot", "static wide"

→

Avoid complex scenes with multiple moving subjects — keep it focused

→

Use for b-roll footage to supplement your main content

→

Storyboard mode lets you create multi-scene sequences