Nano banana face swap prompts are text instructions that tell Gemini-style image models to replace one face with another while preserving lighting, angle, and expression. The key is to describe both the source and target in detail, specify the output format, and avoid vague terms like "swap faces." This guide gives you copy-paste prompts for common scenarios, plus the structure that makes them work.
Why Nano Banana Face Swap Prompts Need More Than "Swap Faces"
A prompt like "swap the face" fails because the model has no anchor for identity, lighting, or pose. Nano banana models generate images from scratch based on your text; they don't have a built-in face-swap tool. You must describe the target face as if you were directing a photo shoot: age, features, expression, and how the light hits the skin. The model then synthesizes a new image that matches your description. For a deeper look at how to control facial features, see ai-portrait-eye-details-control and ai-portrait-skin-texture-makeup.
The Anatomy of a Face Swap Prompt
Every effective nano banana face swap prompt has six parts. Miss one and the result looks pasted or plastic.
- Source description: The original photo's subject, pose, and lighting. Example: "a 30-year-old woman with shoulder-length brown hair, three-quarter view, soft window light from the left."
- Target identity: The face you want to insert. Example: "the face of a 45-year-old man with a short beard, deep-set eyes, and a prominent nose."
- Expression and mood: "neutral expression, slight smile, serious gaze."
- Lighting and environment: "golden hour backlight, warm tones, outdoor cafe background."
- Camera and lens: "85mm lens, f/1.8, shallow depth of field."
- Aspect ratio and output: "3:2 aspect ratio, photorealistic, 4K."
This structure prevents the model from inventing random features. It also lets you reuse the same prompt with different identities. For a full syntax breakdown, see prompt-structure-syntax-guide.
Copy-Paste Prompts for Common Face Swap Scenarios
1. Professional Headshot Swap
Use this when you need a consistent corporate headshot but want to change the person.
"A professional headshot of a 40-year-old woman with a sleek bob haircut, dark brown eyes, and a confident closed-mouth smile. She is wearing a navy blazer over a white shirt. The background is a soft gray gradient. Lighting is a large softbox from the front-left, creating a subtle catchlight in her eyes. Shot with an 85mm lens at f/4, 1:1 aspect ratio, photorealistic, high detail on skin texture and hair strands."
To adapt this for a man, change "woman" to "man," "bob haircut" to "short cropped hair," and "navy blazer" to "charcoal suit." For more headshot-specific tips, see ai-headshot-prompts-guide.
2. Cinematic Film Still Swap
This prompt mimics a movie frame with dramatic lighting.
"Cinematic film still of a 28-year-old man with tousled black hair, a scar above his left eyebrow, and a tense expression. He is standing in a rain-soaked alley at night, illuminated by a single neon sign reflecting off wet pavement. Lighting is high-contrast, with deep shadows and a cool blue-teal color grade. Shot on Arri Alexa with a 40mm anamorphic lens, 2.39:1 aspect ratio, film grain, shallow depth of field."
For more cinematic prompts, see ai-cinematic-film-stills-prompts.
3. Fantasy Character Swap
Use this to place a real face onto a fantasy character.
"A fantasy warrior portrait of a 35-year-old woman with braided silver hair, violet eyes, and a determined expression. She wears ornate leather armor with metal pauldrons. The background is a misty forest at dawn, with soft volumetric light filtering through trees. Lighting is warm and directional from the upper right. Shot with a 50mm lens at f/2.8, 3:2 aspect ratio, hyper-realistic, detailed skin pores and armor scratches."
For more fantasy character design, see ai-fantasy-scifi-character-design.
4. Vintage Photo Swap
This prompt recreates an old photograph with a new face.
"A vintage 1950s portrait of a 50-year-old man with a receding hairline, thick eyebrows, and a gentle smile. He is wearing a tweed jacket and a fedora. The photo has a sepia tone, slight grain, and soft focus. Lighting is a single tungsten lamp from the right, creating a warm glow. Shot with a medium format camera, 4:5 aspect ratio, aged paper texture, slight vignette."
For more vintage styles, see ai-vintage-retro-photography.
5. Social Media Profile Swap
This prompt creates a clean, modern profile picture.
"A close-up portrait of a 22-year-old woman with curly red hair, freckles, and a bright smile. She is looking directly at the camera. The background is a blurred city street at dusk with bokeh lights. Lighting is soft and even, with a slight rim light from behind. Shot with a 35mm lens at f/1.8, 1:1 aspect ratio, vibrant colors, high resolution."
For more portrait lighting tips, see ai-portrait-lighting-techniques-prompts.
Advanced Techniques for Nano Banana 2 Face Swap
Nano Banana 2 models handle more complex prompts. You can specify multiple people, precise expressions, and even subtle skin details. Here are three advanced tactics.
Multi-Person Scene
When swapping faces in a group, describe each person separately and their positions.
"A family dinner scene with three people. On the left, a 60-year-old woman with gray hair in a bun, wearing a floral blouse. In the center, a 35-year-old man with a beard, wearing a blue sweater. On the right, a 10-year-old girl with pigtails, wearing a yellow dress. They are sitting at a wooden table with a roast chicken and candles. Lighting is warm from overhead pendant lamps. Shot with a 24mm lens at f/2.8, 16:9 aspect ratio, candid moment."
For more multi-person composition, see ai-multi-person-scene-composition.
Expression Control
You can dictate the exact emotion. Use words like "micro-smile," "furrowed brow," or "squinting eyes." For a full guide, see portrait-expressions-emotions-control.
Skin Texture and Age
To avoid a plastic look, specify skin details: "visible pores, fine lines around eyes, slight redness on cheeks." For more, see ai-portrait-skin-texture-makeup.
Common Mistakes and How to Fix Them
Mistake 1: Vague identity. "A young man" gives you a random face. Fix: "a 25-year-old man with a square jaw, deep-set brown eyes, and a slight stubble."
Mistake 2: Ignoring lighting. If the source photo has hard shadows, but you describe soft light, the face won't blend. Fix: match the lighting description to the original.
Mistake 3: Forgetting aspect ratio. Without it, the model may crop unexpectedly. Fix: always include "3:2" or "1:1."
Mistake 4: Overloading with details. Too many adjectives can confuse the model. Fix: prioritize the five most important features.
Mistake 5: No negative prompts. Add "no distorted features, no extra limbs, no blurry face" to avoid artifacts. For more, see negative-prompts-guide-ai-images.
Testing and Iteration
Generate four variations at a time. Compare them side by side. If the face looks off, adjust one variable: lighting, expression, or camera angle. Keep a log of what worked. Nano banana models are sensitive to word order; put the most important details first.
Ethical and Practical Considerations
Only swap faces with consent. Do not use real people's faces without permission. For commercial work, ensure you have model releases. Nano banana prompts are for creative and editorial use; respect privacy and copyright.
Try NanoPrompts for Ready-Made Face Swap Prompts
NanoPrompts offers a library of copy-paste prompts for face swaps, headshots, style transfers, and 3D collectible figures. Instead of writing from scratch, browse templates that already include lighting, camera, and aspect ratio details. Visit NanoPrompts to find prompts for Gemini and nano-banana style models. Start with a template, tweak the identity, and generate consistent results in minutes.
