Is this a full video prompt?

This page covers the still keyframe; pair it with your I2V model’s motion prompt for the clip.

How do I keep character consistency?

Reuse the same reference still and face ID notes across rerolls before animating.

chatgpt image prompts

ChatGPT Image anime video prompt

Build a stylized character keyframe first, then animate with your video model using motion notes from the same brief.

Example prompt

Anime cel-shaded heroine, three-quarter view, wind in hair, saturated twilight sky, clean line art, subtle film grain, 16:9 keyframe, consistent face ID seed A17, no text, no logos.

Copy into ChatGPT Image / GPT Image; adjust product names and hex colors.

Steps

Step 1
Generate the keyframe
Keep face angle consistent; save the seed or reference image for the video step.
Step 2
Add motion instructions
Export still, then in your video tool request 3–5s loop: hair + cloth only, camera static.
Step 3
Check loop seams
Trim first/last frames in edit if the loop hiccups; avoid heavy camera pans on cel-shaded art.

Multi-step pipeline: Workflow index · Recipe library

Related prompts

FAQ

Is this a full video prompt?: This page covers the still keyframe; pair it with your I2V model’s motion prompt for the clip.
How do I keep character consistency?: Reuse the same reference still and face ID notes across rerolls before animating.