
These AI prompts create four kinds of visual effect: digital glitch, drifting smoke, surreal photo manipulation and double exposure silhouettes. Some prompts edit a photo you upload, so only use images you have permission to edit. A clear subject and a simple background usually give the best result.
What’s on This Page
| Section | Prompts | Best for |
|---|---|---|
| Glitch effect | 6 | Album art and cyberpunk designs |
| Smoke Effect Prompts | 6 | Adding depth to a photo |
| Surreal Photo Manipulation Prompts | 6 | Dreamlike edits of your images |
| Double Exposure Nature Prompts | 6 | Landscapes inside a silhouette |
| Newborn Photo Prompts | 9 | Enhancing baby photos |
| Antique Portrait Prompt | 1 | A vintage photo look |
| Mixed-Media Doodle Overlay Prompts | 6 | Hand-drawn doodles over photos |
The look began with real errors, such as bad signals and corrupted files, and artists later adopted it on purpose. Now you can create it by describing the damage you want.
Why the Glitch Effect Works
Most digital design is clean and predictable, so distortion catches the eye. The effect suits:
- Album art and cyberpunk designs
- Social media graphics that need movement
- Wallpapers and posters
Keep part of the subject readable. If nothing is recognisable, the image just looks broken.
How to Write a Glitch Prompt
Tools such as Nano Banana and Adobe Firefly respond better to specific damage than to “add glitch”. Useful words include pixel displacement, RGB separation, scan lines and digital noise.
Before you generate, replace these placeholders:
- <SUBJECT> or <UPLOADED IMAGE>: a person, object, logo or text
- <AR>: your frame size, such as 9:16 or 1:1
6 AI Prompts for a Glitch Effect

RGB Channel
Replace the highlighted text with your own details before you run the prompt.

VHS-Style
Replace the highlighted text with your own details before you run the prompt.

Data Corruption
Replace the highlighted text with your own details before you run the prompt.

Minimal Glitch
Replace the highlighted text with your own details before you run the prompt.

Typography
Replace the highlighted text with your own details before you run the prompt.

Extreme Glitch
Replace the highlighted text with your own details before you run the prompt.
Which Style to Start With
Start with Minimal Glitch if you’re using a logo or portrait, since faces and letters stay readable. For wallpapers and album covers, try Data Corruption or Extreme Glitch. If it’s too messy, remove one type of damage from the prompt.
Smoke Effect Prompts
These six AI prompts add a smoke effect to your uploaded photo, from a faint haze to dense fog and swirling energy.
Smoke looks most natural when it sits behind or around the subject and catches the existing light. Start with Minimal Soft Haze if you’re working with a portrait, because heavy smoke can hide the face.

Cinematic Drift Flow
Replace the highlighted text with your own details before you run the prompt.

Dense Fog Compression
Replace the highlighted text with your own details before you run the prompt.

Backlit Glow Smoke
Replace the highlighted text with your own details before you run the prompt.

Swirling Energy Motion
Replace the highlighted text with your own details before you run the prompt.

Minimal Soft Haze
Replace the highlighted text with your own details before you run the prompt.

High Contrast Smoke Separation
Replace the highlighted text with your own details before you run the prompt.
Surreal Photo Manipulation Prompts
These six AI prompts turn an uploaded photo into a surreal photo manipulation, with floating worlds, a tear in reality, reverse gravity and melting scenes.
Upload a clear image, paste a prompt and generate. Results look more believable when the light on the new elements matches the original, so ask for matching shadows if something looks pasted on. Only use photos you’re allowed to edit.

Gravity Collapse Floating World
Replace the highlighted text with your own details before you run the prompt.

Reality Tear Portal Distortion
Replace the highlighted text with your own details before you run the prompt.

Massive Scale Overload Scene
Replace the highlighted text with your own details before you run the prompt.

Reverse Gravity Spatial Shift
Replace the highlighted text with your own details before you run the prompt.

Liquid Reality Melt Transformation
Replace the highlighted text with your own details before you run the prompt.

Infinite Recursive Depth Illusion
Replace the highlighted text with your own details before you run the prompt.
Double Exposure Nature Prompts
These six AI prompts create double exposure nature silhouettes, where a landscape appears inside the shape of a person or animal. They include a wolf filled with mountains, a bear filled with forest and a man’s profile filled with ocean.
A clear outline is key, so choose a subject you can recognise from its shape alone. A plain background helps too. To make your own, keep the “double exposure” wording and change the subject and the scene inside it.

Mountain Wolf Fusion

Blossom Human Portrait

Forest Bear Silhouette

Ocean Man Profile

Horse Landscape Blend

Couple Galaxy Nature Mix
Newborn Photo Prompts
These AI prompts improve a photo you’ve already taken of your baby. They adjust light, colour and clarity, and ask the AI to keep your baby’s face, pose and background as they are.
Most home baby photos have the same problems: uneven light, dull colour and indoor shadows. You don’t need a studio to fix them. Upload your photo to an image tool such as Nano Banana, paste a prompt and compare the result with the original.
Before you upload a photo of your child, read the tool’s privacy settings. Look closely at the result too, because AI can still change small details like eyes, fingers and skin even when told not to. Keep your original.
The One Rule for Every Prompt
Every prompt follows one rule: your original photo stays recognisable.
- No pose changes
- No facial changes
- No new background
- No added props
One Prompt for Every Milestone Month
Use one prompt across the first 12 months, changing only the month number, to give your milestone photos a matching look.
Newborn Skin Tone & Texture Correction
Sleepy Newborn Calm Enhancement
Swaddled Baby Comfort Enhancement
Baby Held in Parent’s Arms
First Smile Preservation
Emotional Warmth Enhancement
Tummy-Time Balance Enhancement
Quiet Play Discovery Moment
Crawling Motion Cleanup
Why This Prompt Keeps Photos Consistent
- Changes stay small, so the baby remains the focus.
- There are no bright or busy elements.
- You only change one number for each month.
How to Know When to Stop
A good edit is one you don’t notice. If you see the editing before you see the baby, use a gentler prompt or go back to the original.
Antique Portrait Prompt
This AI prompt turns a portrait photo into an antique portrait with a vintage engraved look, giving it the faded tones and aged paper of a very old picture.
Use a clear photo with one face and even light. Only upload photos of yourself, or of people who have agreed.
Upload the photo to ChatGPT or Nano Banana, paste the prompt and generate. Compare the result with your original, because AI can change facial features. If it no longer looks like the person, try again and ask it to keep the face unchanged.
Antique Portrait Engraving Prompt
Replace the highlighted text with your own details before you run the prompt.
Mixed-Media Doodle Overlay Prompts
I really like the way Mixedmedia Doodle Photo Overlays can make an ordinary photo feel more personal. The photo stays realistic, while the handwritten marks, little sketches, arrows, notes, and playful shapes make it feel as if someone has taken a pen to the image afterward.
That balance is what makes this style work for me. If the doodles become too perfect, the design starts looking like a regular digital graphic. If there are too many of them, the photo gets lost. The better results usually come from keeping the photograph natural and letting the hand-drawn elements interact with it.
For the images in this article, I used ChatGPT’s image-generation tools. You can also try these ideas with other image-generation tools such as Midjourney, Gemini, Firefly, Ideogram, Leonardo, or Stable Diffusion. The results will not be identical across platforms, but the same creative direction can be adapted to whichever tool you prefer.
I also kept the prompts fairly detailed because small things—such as the texture of the lines, where the doodles sit, the amount of empty space, and how they interact with the person—can make a noticeable difference.Mixedmedia Doodle Photo Overlays AI Prompts
Coquette Dreams Soft Editorial Style
Replace the highlighted text with your own details before you run the prompt.
Master Your Day Productivity Editorial Poster
Replace the highlighted text with your own details before you run the prompt.
Future Astronaut Cinematic Imagination Style
Replace the highlighted text with your own details before you run the prompt.
City Explorer Travel Creator Poster
Replace the highlighted text with your own details before you run the prompt.
Find Your Zen Minimal Wellness Editorial Style
Replace the highlighted text with your own details before you run the prompt.
Leveling Up Dark Gaming Editorial Poster
Replace the highlighted text with your own details before you run the prompt.
Which AI Tool Did I Use?
I used ChatGPT’s image-generation tools to create the images shown in this article.
You don’t have to use ChatGPT to experiment with the prompts. You can try the same creative direction in Midjourney, Gemini, Adobe Firefly, Ideogram, Leonardo, Stable Diffusion, or another image generator you already use.
Just keep in mind that different models interpret the same prompt differently. One might give you better photographic realism, while another may handle the hand-drawn elements or composition in a way you prefer.More in AI Photo Editing Prompts and Guides
This guide is part of our AI Photo Editing Prompts and Guides hub, where you’ll find every related guide, along with tips and common questions.

