BackBlog
General

Nano Banana Pro Prompts for Consistent Makeup Looks (No Face Drift)

Tired of AI-generated makeup looks that change your model’s face shape, skin tone, or features with every new prompt? This guide shares Nano Banana Pro prompt templates, reference photo prep tips, and troubleshooting fixes to create consistent full-face makeup variations while preserving model identity. Perfect for beauty brands, makeup artists, and content creators building cohesive lookbooks or campaign assets.

Nano Banana Pro Prompts for Consistent Makeup Looks (No Face Drift)
MagicBoat AI8 min read72 views
Table of contents

The problem with AI-generated makeup look sets

If you’ve ever tried to build a makeup lookbook or campaign asset set with AI, you’ve likely run into the same frustrating issue: every new prompt produces a different face. One look has a round jawline, the next has high cheekbones, a third has a completely different skin tone—even when you use the same base description. For beauty brands, makeup artists, and content creators, this face drift makes AI-generated makeup sets useless for cohesive campaigns, lookbooks, or shade range showcases.

The good news is that with the right reference setup and prompt structure, you can lock in a model’s identity while swapping out full-face makeup looks. nano banana pro excels at reference-based image generation, making it ideal for creating consistent makeup variations without losing facial features, skin tone, or bone structure.

Prep your reference photo for consistent identity

Your reference photo is the foundation of consistent makeup generation. A low-quality or poorly framed reference will almost always lead to face drift, no matter how good your prompt is. Follow these guidelines to set yourself up for success:

  • Choose a front-facing, eye-level portrait of the model with a neutral expression. Avoid tilted heads or extreme angles, as these make it harder for the AI to preserve facial structure.
  • Use minimal or no makeup in the reference photo. Heavy foundation, bold eyeshadow, or bright lipstick will interfere with the AI’s ability to apply new makeup on top of the skin.
  • Prioritize even, soft lighting (natural window light works best) with no harsh shadows on the face. Uneven lighting or strong color casts can cause skin tone shifts in generated looks.
  • Use a high-resolution image (at least 1024x1024) with clear skin texture. Blurry or low-res references lead to more drift and less realistic makeup application.
  • Crop the image to focus on the face and upper shoulders, with a plain neutral background (white, beige, or light gray). Distracting background elements can pull the AI’s focus away from facial consistency.
Pro tip: If you don’t have a real model photo, generate one base portrait first with a detailed prompt, then use that as your reference for all subsequent looks. This gives you a fully original, on-brand model to build your entire look set around.

Core prompt framework for no-face-drift makeup looks

A well-structured prompt balances creative freedom for makeup design with strict guardrails for identity preservation. The most consistent results come from a 4-part framework that tells the AI exactly what to keep the same and what to change:

  1. Reference anchor: Explicitly instruct the AI to use the reference photo for all facial features, and set a reference weight high enough to preserve identity (0.8–0.9 works for most use cases).
  2. Makeup look details: Specify every element of the makeup in granular detail—base coverage and finish, blush shade and placement, eyeshadow shades, eyeliner style, lip finish, and more. Vague descriptions lead to random AI choices and subtle drift.
  3. Lighting and setting: Match the lighting and background across all prompts to keep the set cohesive. Inconsistent lighting is a common cause of perceived skin tone shifts between looks.
  4. Consistency guardrails: Add explicit negative and positive prompts to prevent unwanted changes to face shape, skin tone, eye color, age, or facial features.

Use this core template as a starting point for all your makeup look prompts:

Use the reference image for exact facial structure, skin tone, eye color, and all facial features. Reference image weight 0.85.
Full-face makeup look: [specific makeup details, e.g., dewy light-medium coverage foundation, peachy pink cream blush on the apples of the cheeks, soft bronze contour along the jawline, champagne highlighter on the high points, warm taupe eyeshadow blended into the crease, thin black liquid eyeliner, brown mascara, groomed brows, peachy nude satin lipstick]
Lighting: soft natural window light, even illumination, no harsh shadows
Background: plain warm beige studio background
Photorealistic, high resolution, beauty editorial quality, sharp focus on the face
Negative prompt: face shape change, skin tone shift, eye color change, age change, distorted features, blurry, grainy, heavy filters

Want to test the framework with your own reference? try this prompt for free to generate your first consistent makeup look in minutes.

4 ready-to-use makeup look prompt templates

These pre-built prompts are designed for a 4-look campaign for a multi-use face, eye, and lip collection. All use the same reference anchor and guardrails, so you can generate them back-to-back with the same base photo for a fully cohesive set.

Everyday natural look

Use the reference image for exact facial structure, skin tone, eye color, and all facial features. Reference image weight 0.85.
Full-face makeup look: sheer light-reflecting foundation with natural finish, subtle cream blush in soft rose blended high on the cheeks, light bronzer along the hairline, champagne highlighter with a sheer finish on the cheekbones, neutral taupe eyeshadow swept lightly across the lids, clear brow gel, brown mascara on top lashes only, tinted lip balm in a muted rose shade
Lighting: soft natural window light, even illumination, no harsh shadows
Background: plain warm beige studio background
Photorealistic, high resolution, beauty editorial quality, sharp focus on the face, natural skin texture
Negative prompt: face shape change, skin tone shift, eye color change, age change, heavy makeup, distorted features, blurry, grainy, filters

Office polished look

Use the reference image for exact facial structure, skin tone, eye color, and all facial features. Reference image weight 0.85.
Full-face makeup look: medium coverage satin foundation, soft terracotta powder blush blended on the apples of the cheeks, defined contour along the jawline and cheekbones, subtle champagne highlighter, warm brown eyeshadow in the crease with a matte ivory lid, thin brown eyeliner tightlined on the upper lash line, black mascara on top and bottom lashes, groomed filled brows, creamy mauve lipstick with a satin finish
Lighting: soft studio ring light, even bright illumination, no harsh shadows
Background: plain off-white studio background
Photorealistic, high resolution, professional beauty headshot, sharp focus on the face, smooth natural skin texture
Negative prompt: face shape change, skin tone shift, eye color change, age change, glitter, bold makeup, distorted features, blurry, grainy

Date night glam look

Use the reference image for exact facial structure, skin tone, eye color, and all facial features. Reference image weight 0.85.
Full-face makeup look: full coverage dewy foundation, deep rose cream blush blended on the cheeks, sculpted bronzer, intense champagne highlighter on the high points of the face, warm bronze eyeshadow with a shimmery gold lid, winged black liquid eyeliner, false lash effect mascara, filled defined brows, deep cherry red lipstick with a glossy finish
Lighting: warm soft studio lighting, subtle glow, gentle shadow under the jawline
Background: plain deep burgundy studio background
Photorealistic, high resolution, beauty campaign quality, sharp focus on the face, dewy skin texture
Negative prompt: face shape change, skin tone shift, eye color change, age change, distorted features, blurry, grainy, matte skin

Festival bold look

Use the reference image for exact facial structure, skin tone, eye color, and all facial features. Reference image weight 0.85.
Full-face makeup look: medium coverage dewy foundation, bright tangerine cream blush blended high on the cheeks and temples, bronze contour, iridescent pink highlighter on cheekbones, brow bone, and cupid's bow, bright electric blue eyeshadow on the lid with a warm orange transition shade, glitter liner along the upper lash line, volumizing black mascara, feathered filled brows, bright orange matte lipstick with a clear gloss top coat
Lighting: bright natural sunlight, warm golden hour tone
Background: plain pale yellow studio background
Photorealistic, high resolution, bold beauty editorial, sharp focus on the face, dewy glowing skin texture
Negative prompt: face shape change, skin tone shift, eye color change, age change, distorted features, blurry, grainy, muted colors

Troubleshooting common face drift issues

Even with the perfect reference and prompt, you may run into small inconsistencies. Here’s how to fix the most common face drift problems:

  • Face shape or jawline changes: Increase the reference image weight by 0.05–0.1 (up to 0.95) and add "preserve exact jawline and face shape from reference image" to your prompt. Avoid heavy contour or makeup that covers the jawline, as this can confuse the AI.
  • Skin tone shifts: Double-check that your reference photo has even, neutral lighting with no warm or cool color casts. Add "exact skin tone match to reference image, no color shift" to your prompt, and make sure your lighting description matches the reference’s lighting temperature.
  • Eye color or eye shape changes: Add "preserve exact eye color and eye shape from reference image" to your prompt. If you’re using bold eyeshadow, specify that the eye shape itself should not change—only the makeup applied to the lids.
  • Age or texture changes: If the AI is making the model look older or younger, add "preserve exact age and skin texture from reference image" to your prompt. Avoid descriptive words like "youthful" or "mature" unless you intend to shift the model’s apparent age.
  • Background color bleed: If the background color is tinting the model’s skin or hair, add "no color cast from background, natural skin tone" to your negative prompt, or use a neutral background that matches the reference’s background tone.

Final checklist

  • Use a high-res, front-facing reference photo with minimal makeup and even, neutral lighting
  • Include a clear reference anchor with a weight between 0.8 and 0.9 for most use cases
  • List every makeup detail explicitly to avoid random AI choices and inconsistent looks
  • Add consistency guardrails for face shape, skin tone, eye color, and age to prevent drift
  • Match lighting and background across all prompts for a cohesive, professional look set
  • Test one look first and adjust reference weight before generating the full set of looks
  • Use an ai image generator with strong reference image control for the most consistent results

Keywords

ai makeup generator, nano banana pro, makeup look generator, consistent ai makeup, face drift ai, beauty campaign ai, full face makeup prompts