Have you ever seen a cool, crazy, or beautiful picture online and found out it was made by a computer? That’s the magic of AI, or Artificial Intelligence. AI is like a smart computer program that can learn and create things. One of the most fun things AI can do is create images from text. This text is called an “image prompt.”
An image prompt is like a set of instructions you give to an AI. The more clear and detailed your instructions are, the better and more accurate the final picture will be. It’s like telling a painter exactly what you want to see in your painting.
There are many AI tools available today. Some are for chatting, like ChatGPT and Google Gemini, which can also help you write good image prompts. Others are specially made for creating images and videos. Popular image-making tools include Midjourney, DALL-E, and Stable Diffusion. Google also has its own AI tool for making images, which is called Gemini (formerly known as Bard).
This article will teach you how to use a detailed image prompt to create your own picture. We will use Google Gemini for this. The special prompt we are using will create a hyper-realistic portrait with fire and water effects. Remember, the key is to be very specific.
Below is the exact image prompt you will need to use. You must copy it exactly as it is written for it to work properly.
The Exact Image Prompt to Use
Please copy the entire text inside the box below. Do not change any words.
Create a hyper-realistic DSLR-quality portrait using the attached PERSON image only for facial likeness. Use only the reference image head and neck down to the collarbone. Frame tightly so the crop stops exactly at the collarbone, with nothing below included. The portrait must show bare skin only on the neck and collarbone area, no clothing or fabric of any kind. The subject is upright, head tilted slightly downward so the chin is near the chest. Eyes must look directly into the camera lens. Expression: serious with a subtle smirk, showing a faint smile only on the right corner of the mouth. Wet hair framing the face naturally, hairstyle must match the reference image closely in shape and flow.
character_likeness: {
"use_attached_image": true,
"lock_identity": true,
"reference_weight": 2.0,
"consistency": "Preserve the exact likeness of the reference face, including bone structure, proportions, skin tone, hairstyle, and collarbone area. Hair must match the reference image in direction, wetness, and flow. Only head and neck down to collarbone may be used. Crop must stop exactly at the collarbone. Do not generate chest, torso, or any body below collarbone. Do not auto-generate tattoos. Only replicate tattoos if they are clearly visible in the reference image.",
"negative_prompt": "no chest, no torso, no armpit, no nipples, no zoomed out crop, no clothing, no t-shirt, no fabric, no straps, no garments, no invented tattoos, no extra body length below collarbone"
}
composition: {
"framing": "tight portrait crop head and neck down to collarbone only, head dominant",
"pose": "upright posture, head slightly tilted downward, chin near chest",
"gaze": "eyes locked directly into lens",
"expression": "serious with subtle smirk, slight smile on right corner of mouth",
"shoulders": "bare skin neck and collarbone only, cropped at collarbone",
"hair": "must replicate wet hairstyle shape and flow exactly as in reference"
}
scene: {
"background": {
"left_side": {
"color": "deep fiery orange with black smoke",
"fire_origin": "erupting directly from cheek, temple, and wrapping across left side of neck and collarbone",
"fire_style": "dense flames wrapping both behind and in front of the subject, licking across skin on the left side of face, neck, and collarbone, flowing upward with force",
"embers": "super dense glowing embers saturating left half, layered both in front of and behind subject, reflecting off wet hair and skin"
},
"right_side": {
"color": "steel blue and black haze",
"water_origin": "continuous waterfall curtain pouring down from the very top edge of the frame, saturating the entire right half",
"water_style": "solid, heavy torrents with maximum density, cascading over hair, face, neck, and collarbone, wrapping tightly around from the front as well as behind",
"mist": "powerful spray and vapor exploding outward across right half, surrounding collarbone and merging into background"
}
}
}
lighting: {
"contrast": "dark cinematic mysterious with maximum intensity across all outputs",
"left": "intense firelight flooding left cheek, jawline, and collarbone, wrapping across neck and glowing directly on skin from the front",
"right": "cold electric blue light reflecting heavily across waterfall streams, droplets, right cheek, jawline, neck, and collarbone",
"front": "sharp high contrast key light defining facial structure and wet skin with cinematic intensity",
"skin": "wet reflective sheen across face, neck, and collarbone glowing with fire on left and water on right, both elements visibly wrapping around and engulfing from the front"
}
style: {
"photography": {
"camera": "DSLR",
"lens": "85mm portrait prime",
"aperture": "f/2.0",
"shutter": "fast shutter freezing flames, embers, torrents, and droplets",
"iso": "low for cinematic blacks",
"quality": "hyper realistic 4K+ detail"
},
"mood": "intense dark and cinematic. Fire erupting from the subject’s left side wrapping across the face, neck, and collarbone. Water flooding from the top of the frame and drowning the right side, engulfing the body. Subject calm but with a faint smirk of resilience"
}
effects: {
"gaze": "eyes locked into lens, overriding original gaze direction",
"fire": {
"direction": "flames rising vertically in dense sheets, flowing upward and wrapping forward",
"coverage": "fire fully engulfing left face, neck, and collarbone from the front and sides, merging into background",
"embers": "extremely dense arcs scattering upward and outward across left half"
},
"water": {
"streams": "solid waterfall curtain pouring from the very top of the frame, unbroken and continuous",
"coverage": "water cascading directly over right side of face, hair, neck, and collarbone, wrapping fully around the subject from the front and sides",
"droplets": "dense beads shimmering in icy blue light, suspended in midair across skin",
"spray": "explosive mist cloud engulfing right side and collarbone from all directions"
},
"atmospherics": {
"smoke": "thick black orange smoke interlaced with flames wrapping tightly around the left side",
"mist": "icy blue haze saturating right side and wrapping around body, heavy density"
}
}
output: {
"format": "high resolution PNG",
"dpi": 300,
"consistency": "Portrait optimized for social platforms. Only head and neck down to the collarbone may be used. Crop exactly at collarbone, with no armpit or chest visible. Bare skin only, no clothing or fabric. Do not auto-generate tattoos. Only replicate tattoos that are actually visible in the reference image. Hairstyle must match reference. Eyes locked into lens. Expression shows subtle smirk with slight smile on right corner of mouth. Fire must fully wrap around face, neck, and collarbone on the left. Water must pour from the very top of the frame, cascading heavily across right side of face, hair, neck, and collarbone, engulfing from the front and sides. Lighting must always remain dark cinematic and mysterious across all outputs.",
"credit": "Nano-Banana Prompt Engineering"
}
How to Create Your Image in Google Gemini
Now, follow these simple steps to turn this prompt into your own image.
Step 1: Go to Google Gemini
Open your web browser and go to the Google Gemini website (https://gemini.google.com). Make sure you are logged into your Google account.
Step 2: Upload Your Reference Photo
This prompt needs a reference photo to copy the person’s face. Look for the image upload icon (usually a little picture or a ‘+’ sign) in the chat box. Click on it and select a clear, front-facing photo of the person you want to be in the portrait. This is very important.
Step 3: Paste the Prompt
Click on the text input field in Gemini. Paste the entire prompt you copied above. Be careful not to change or miss any part of it.
Step 4: Generate the Image
Press ‘Enter’ or click the send button. Google Gemini will now read your instructions and the reference photo and start generating the image. This may take a minute or two.
Step 5: Review and Download
Once the image is generated, look at it carefully. Check if it followed all your instructions, like the fire on the left and water on the right. If you are happy with the result, you can usually download it by clicking on the image and looking for a download button.
And that’s it! You just used a powerful AI tool to create a professional-looking, creative image from a text prompt. Remember, practice makes perfect. Try changing small details in the prompt to see how it changes your image. Happy creating