Realistic gym mirror‑selfie of ME in a modern fitn...

Tested & Verified Nano Banana Commercial License

Basic Prompt

Realistic gym mirror selfie of a fit woman in a modern fitness center. Three-quarter body shot from a side-back angle: she holds a smartphone in her right hand, glancing toward the mirror with a relaxed, confident expression. Outfit: light neutral sports bra and high-waisted beige seamless workout shorts; sleek smartwatch on left wrist; tiny gold stud earrings. Lighting: bright indoor gym lighting with balanced cool and neutral tones clean highlights on shoulders and arms, soft shadows defining natural muscle tone, no harsh contrast. Background: lively gym environment with blurred figures training, weight benches, dumbbell racks, and cardio machines on rubber flooring. A bold yellow accent wall/column appears on the right edge of the frame. Mirror reflection reveals depth more equipment, ceiling lights, distant treadmills. Slight motion blur and background bokeh keep focus on the subject. Color palette: gym neutrals (black, gray, steel) + warm beige outfit + vibrant yellow wall. Natural skin tones, photorealistic detail, 4K resolution.

To view Detailed Prompt please login

Add to Library

This image captures authentic fitness culture not perfection, but presence.

No approved reviews yet.

Write a comment

Select a rating

Model Settings & Parameters

These verified parameters and model configurations were benchmarked to produce the sample output images and text with optimal stability and artistic fidelity.

Target Engine Nano Banana

Aspect Ratio 16:9 / 1:1

Stylize / CFG 250 – 750

Output Resolution UHD 4K / 8K

Seed Stability Random / Multi-run

Generation Time < 30 seconds

Known Limitations & Tips

Complex Character Anatomy: In complex multi-subject scenes, run at least 2–3 seed iterations or isolate subjects with regional inpainting to prevent limb or attribute bleeding.
Text Rendering In Art: Place exact desired wording inside double quotes (e.g. "Neon Sign"). Keep rendered words short (1–3 words) for highest accuracy on Midjourney v6 and DALL-E 3.
Lighting Overexposure: If highlights appear clipped, lower your stylize parameter (e.g. --s 150) or append diffused volumetric ambient lighting.