Pick your fighter. Pick the fit. Pick the scene. Throw it in Grok Imagine. 15 seconds of glory.
① Select Character
ECHO
The Nun · White Veil
MATHILDA
Beanie + Round Sunglasses
② Select Costume
③ Select Scenario
Your Grok Imagine prompt appears here…
Attach the costume's headshot ref in Grok Imagine. Prompt describes what's NEW — Grok reads the face from your ref.
💡 GROK IMAGINE PROMPTING TIPS (researched)
Lead with the subject. The first words carry the most weight — Grok commits hardest to what it reads first. Put character + action up front.
Write like a director, not a typist. Full sentences, not keyword soup. "She strides through the neon market, coat catching the wind" beats "woman, neon, market, coat".
Light like it matters — because it does. Lighting is the single highest-leverage detail. Name it: "golden hour backlight", "hard rim light from the left", "overcast diffused".
Name a style early. "Cinematic film still", "photorealistic", "Blade Runner mood" gives Grok a visual library. Vague adjectives ("dark", "cool") don't.
One focus per prompt. Crowded prompts with competing subjects produce mush. 30–80 words is the sweet spot.
Positive direction only. Describe what you want, never what to avoid. (Grok's filter reads the words either way.)
Camera language works. "Low-angle", "85mm portrait lens", "shallow depth of field", "slow dolly forward" — Grok understands real photography terms.
For video: give motion a verb. "The coat lags half a beat behind each turn" beats "coat moves". Go wider for people — wide shots + slow moves are cleanest.
Reference images: prompt what's NEW. When you attach a ref, don't re-describe the face — describe the pose, setting, and costume change. Grok keeps the face from the photo.
15-second Bad Baddies formula: [Character + costume + action] → [environment] → [lighting] → [camera move] → [mood]. Keep it tight, one beat per clip.