Image models

GPT Image

Best for detailed storyboard briefs, title graphics and visual revisions.

Start creating

GPT Image is OpenAI’s image generation family. Prompt following and image editing make it useful for storyboards, pitch frames and production graphics that need a precise brief or visible text.

OpenAI Sunburst gallery example: a cinematic location concept.
Official model example

OpenAI Sunburst gallery example: a cinematic location concept.

View source

GPT Image 2.5 Sunburst

Prioritizes precise image editing. Choose it when an approved frame needs a prop, costume or background change without rebuilding the scene.

Official source

GPT Image 2.5 Flare

The faster everyday image option. Use it to explore several pitch frames or scene concepts before committing to detailed edits.

Official source
01

Describe composition, lighting and props in one detailed shot brief.

02

Iterate on an existing image instead of recreating a scene from scratch.

03

Prototype graphics and signs whose wording matters to the story.

Write the shot intent and required objects, generate several frames, then revise the chosen image. Use approved frames in the pitch deck or as references for video shots.

Before final delivery

Proofread all visible text and inspect object placement, continuity and rights before publishing. Generated graphics still need human approval.