You have seen them on your feed. The cinematic collage portrait. The childhood-meets-present split photo. The Spiderman edit. The scrapbook aesthetic. The flash filter that makes you look like a paparazzi shot. Everyone is asking the same thing: what is the prompt?
Here are 50+ of them. Copy, paste, generate. I update this list every time a new trend hits.
And yes, there is a section at the end explaining why your AI photo looks fake and exactly what to do about it.
Before we get into the prompts, one thing.
If you want to go deeper than just copy-pasting prompts, I put together a full guide with 500+ curated image styles and prompts tested across Midjourney, DALL-E, Stable Diffusion, Nano Banana, Leonardo and more. Every style is organised by category so you can find the exact look you want in seconds.
Photography styles. Cinematic. Fine art. Fantasy. 3D. All of it. All copy-paste ready.
Get it here: AI Image Generation Guide (500+ Styles)
It is the reference I go back to every time I need to generate something that actually looks good. Worth having.
What Are AI Photo Editing Prompts?
An AI photo editing prompt is a text instruction you give to an AI tool like ChatGPT, Gemini, or Midjourney to generate or transform an image. The prompt describes the subject, lighting, camera style, mood, color palette, and visual effect you want.
The difference between a mediocre AI image and a viral one is almost always the prompt. A weak prompt gives you something average. A specific, structured prompt pulls the model toward a precise visual result.
This guide gives you the specific ones.
Which AI Tool Should You Use?
Before we get into the prompts, a quick note on tools because the same prompt produces very different results depending on where you paste it.
ChatGPT (GPT-4o image generation): Best for identity accuracy. When you upload a photo of yourself and want the AI to keep your face recognisable, ChatGPT handles this better than most tools right now. Use it for collage prompts and the childhood vs present trend.
Gemini (Google): Consistently good with South Asian skin tones, traditional outfits, and cultural aesthetics. If you are creating content with an Indian aesthetic, Gemini often reads the cultural context better than competitors.
Midjourney: The strongest tool for cinematic and artistic quality. Fantasy edits, dramatic lighting, fine art styles. The output ceiling is higher here than anywhere else. The tradeoff is that following very literal instructions is not its strength.
Flux (via Leonardo or Replicate): The most photorealistic skin and environment rendering available right now. Use it for portrait edits where you want the result to look like a real photograph.
Ideogram: Use this any time your image needs readable text. Street signs, poster typography, captions in the image itself. The other tools still struggle with this.
Childhood vs Present Trend
This is the biggest AI photo trend on Instagram right now. Users are recreating side-by-side portraits of their younger self and current self in a cinematic, emotional style.
Prompt 1:
On the left: a child version of me from a childhood photo, looking with an innocent smile to the right, with "2007" written above. On the right: a present version of me from a recent photo, sitting with hands under chin, looking at the child with a calm smile, with "2026" written above. In the middle: a simple elegant table. Background: soft studio with warm lighting. Style: professional, minimal, emotional photography focusing on visual communication between the two versions. Make it look like a real photoshoot. Keep my original features without changing. Photo size 4:5.Best tool: ChatGPT (upload both photos)
Prompt 2:
Split screen portrait, left side shows a young child version of me aged 5 to 8, right side shows current adult me, both looking toward each other, warm golden studio lighting, soft shadow, cinematic color grade, nostalgic mood, photorealistic, 4:5 ratioBest tool: ChatGPT or Gemini
AI Collage Prompts
The aesthetic editorial collage is dominating creator feeds. Multiple cropped frames of the same face, flowers, music overlays, warm tones.
Prompt 3 (Aesthetic Portrait Collage):
A warm-toned cinematic portrait collage. Composition: aesthetic editorial collage layout with multiple cropped frames of the same face, close-ups of eyes, lips, and side profile, all consistently matching the reference identity. Maintain high facial consistency across all frames. Add soft floral elements like beige and orange lilies arranged around the frame. Include a subtle blurred music player overlay with a nostalgic vibe, minimal and not distracting. Style: editorial, aesthetic Instagram collage, warm color grading, cinematic lighting, shallow depth of field, ultra-realistic, 4K detail. Important: use the uploaded image as the primary facial reference. Ensure high identity fidelity and consistency. Avoid face distortion, over-smoothing, or AI-style alterations. Keep features natural and recognisable.Best tool: ChatGPT
Prompt 4 (Moody Scrapbook):
A warm moody scrapbook-style portrait collage of the person in the uploaded photo. Multiple overlapping printed portrait photos of the same person including front view, side profile, and extreme eye close-up. Add torn paper, masking tape, vintage floral stickers, aged notebook scraps, and a retro music card detail. Amber low light, flash-photo look, deep shadows, subtle grain, brown-orange vintage color grading, intimate indie Tumblr moodboard aesthetic.Best tool: ChatGPT
Prompt 5 (Vintage Magazine):
vintage magazine collage, retro fashion editorial, sepia tones mixed with muted colors, torn paper edges, analog textures, classic editorial layout, soft grain, nostalgic lighting, layered composition, ultra detailedBest tool: Midjourney or ChatGPT
Prompt 6 (Golden Night Collage):
Portrait collage, golden night theme, glowing city lights background, luxury jewelry fragments, layered magazine cutouts, cinematic lighting, deep shadows, gold and black color palette, surreal reflections, high fashion editorial style, ultra detailed, soft glow, 4KBest tool: Midjourney
Flash Filter Prompts
The Y2K direct flash aesthetic is everywhere on Reels right now. High contrast, glossy skin, night party energy.
Prompt 7:
young woman at night party, direct camera flash, overexposed highlights, shiny skin, dark background, candid pose, 2000s Instagram flash filter effect, high contrast, grainy texturePrompt 8:
mirror selfie, phone flash visible, reflective glare, bathroom setting, soft messy background, flash filter Instagram style, sharp contrast, influencer vibePrompt 9:
urban street at night, subject walking, harsh flash lighting, deep shadows, paparazzi style, motion blur, Instagram flash effect, cinematic realism, cool tonesPrompt 10:
crowded nightclub, colorful lights, strong flash overpowering ambient lighting, glowing skin, flash filter effect, chaotic candid energy, high ISO grainBest tool for all flash prompts: Flux or ChatGPT
Spiderman AI Photo Prompts
Prompt 11 (Male):
A realistic high-angle full-body shot of a young man with messy textured black hair looking up at the camera. He is dressed in a red and blue Spider-Man bodysuit on his upper half but wearing loose casual grey sweatpants on his lower half, white clean canvas shoes. He has a black backpack on his shoulders and is holding a red Spider-Man mask in his left hand. Standing indoors on a grey marble tiled floor. Lighting is soft and natural, giving a casual "Peter Parker after school" vibe.Prompt 12 (Female):
A realistic high-angle full-body shot of a young woman with messy textured black hair looking up at the camera. She is dressed in a red and blue Spider-Man bodysuit on her upper half but wearing loose casual grey sweatpants on her lower half, white clean canvas shoes. She has a black backpack on her shoulders and is holding a red Spider-Man mask in her left hand. Standing indoors on a grey marble tiled floor. Lighting is soft and natural, giving a casual "Peter Parker after school" vibe.Prompt 13 (Cinematic):
Spiderman swinging through New York at sunset, cinematic movie poster style, orange and blue color grade, motion blur on web, city skyline below, dramatic scale, 16:9Best tool: ChatGPT for the portrait versions, Midjourney for the cinematic version
Prompts for Boys (Indian Edition)
South Asian creators, these are built for you. Gemini reads this cultural context well.
Prompt 14:
A fashionable young Indian man sitting relaxed on a rooftop terrace at golden hour, directly facing the camera with a confident slight smile. He wears a white linen kurta with rolled sleeves, slim dark jeans, white sneakers. Background: Mumbai or Delhi city skyline with warm orange haze. Lighting: bright natural sunlight creating a fresh summer mood. Cinematic, high saturation but natural tone, photorealistic, 4:5 ratio.Prompt 15:
Young Indian man in streetwear, graphic tee, joggers, white sneakers, Old Delhi street background with colorful buildings, bokeh background, golden hour, confident attitude pose, editorial photography style, ultra HDPrompt 16:
Indian man in a black bandhgala suit, contemporary wedding aesthetic, outdoor palace garden, soft warm lighting, regal but modern, editorial portrait, Vogue India style, shallow depth of fieldPrompts for Girls (Aesthetic and Soft Glam)
Prompt 17:
A warm-toned cinematic portrait of a young woman with soft glowing skin, subtle freckles, glossy nude lips, long dark brown hair in loose waves. Traditional gold jhumka earrings, delicate necklace, deep maroon outfit with sheer dupatta. Lighting: golden hour inspired with rich amber highlights and soft shadows, cozy intimate mood. Style: soft glam, editorial, aesthetic Instagram, warm color grading, high detail, shallow depth of field, cinematic, 4K.Prompt 18:
Korean aesthetic portrait, young woman, soft diffused window light, dewy skin, minimal natural makeup, white linen background, soft girl aesthetic, calm expression, Fujifilm film grain, 4:5 ratioPrompt 19:
Female portrait, Bollywood editorial style, dramatic studio lighting, jewellery-heavy styling, deep jewel-tone lehenga, smoky eye makeup, marble backdrop, ultra-realistic, high fashion, Vogue India aestheticPrompt 20:
Aesthetic study corner portrait, girl at wooden desk, warm lamp light, rain on window behind her, soft sweater, books and coffee cup on desk, lo-fi cozy mood, warm amber tones, candid editorial feelCouple Prompts
Prompt 21:
Cinematic couple portrait, traditional outfits, coordinated maroon and gold, warm bokeh garden background, soft natural light, faces close together, emotional connection, love story feel, photorealistic, 4:5Prompt 22:
Couple walking on cobblestone street at dusk, city lights blurring behind them, back-lit silhouette with golden halo effect, romantic mood, film photography grain, 16:9Prompt 23:
Photo booth strip style, couple making funny faces and sweet poses, glossy print look with slight curl, doodled hearts around the frame, retro flash look, dark background, high contrast, Fujifilm nostalgic vibeCinematic Portrait Prompts
These work for anyone wanting a professional editorial look without a studio.
Prompt 24:
Transform this portrait into a cinematic night scene with neon lights, rain reflections on the street below, soft glow around face, high contrast shadows, teal and orange color grading, ultra HD quality, shallow depth of fieldPrompt 25:
Studio portrait with professional softbox lighting, plain light gray background, sharp focus, high resolution, clean skin retouching, LinkedIn professional headshot qualityPrompt 26:
Add golden hour lighting with warm sunlight, soft shadows, glowing natural skin tones, subtle hair highlights, outdoor garden setting behind subject, bokeh depth, photorealisticPrompt 27:
Convert this photo into a dramatic fantasy portrait, glowing particle effects around face, dark moody background, ethereal lighting, mystical atmosphere, fine art photography quality, 4:5Color Analysis Prompts
These are trending for a different reason: they are useful, not just aesthetic. People want to know what colors suit them.
Prompt 28:
Analyse the color season of the person in this photo. Determine whether they are Spring, Summer, Autumn, or Winter based on skin undertone, eye color, and natural hair color. Show a recommended color palette of 12 colors that suit their coloring. Present results as a styled editorial card layout.Prompt 29:
Create a personal color analysis chart for this person. Show their season type, best clothing colors, metals that suit them (gold vs silver), and colors to avoid. Clean minimal infographic style with color swatches.Hairstyle and Makeover Prompts
Prompt 30:
Show this person with 4 different trendy hairstyles for 2026: curtain bangs with layers, blunt lob, face-framing highlights with waves, and a sleek high bun. Keep the face completely recognisable. Grid layout, clean studio lighting, ultra-realistic.Prompt 31:
Transform this person with a full makeover: new hairstyle, professional makeup, and a stylish outfit change. Keep original face structure completely intact. Style: editorial fashion shoot, warm studio lighting, 4:5 ratio.Why Your AI Photo Looks Fake (And How to Fix It)
This is the section nobody else has written. You have tried a viral prompt, got a result, and it looked like a wax figure wearing your face. Here is what went wrong and what to do.
Problem 1: The face is distorted or melting
This happens when the AI does not have a clear reference to work from. The fix is adding this line to any portrait prompt: "Keep the face completely natural and recognisable. Avoid distortion, over-smoothing, or AI-style alterations. Use the uploaded photo as the primary facial reference."
In ChatGPT specifically, you need to upload your photo in the same message as the prompt. Describing yourself without a photo reference will always produce a generic face.
Problem 2: The skin tone is completely wrong
AI tools trained on Western datasets default to lighter skin tones unless you specify. Add your skin tone explicitly: "warm medium brown skin tone" or "deep brown skin with warm undertones" or "South Asian golden brown complexion". Do not leave it to assumption.
Problem 3: The style looks generic, not like the trend you saw
You are using a vague prompt. "Make it aesthetic" means nothing to an AI. "Soft warm amber lighting, shallow depth of field, editorial collage layout with floral elements, 4K ultra-realistic" means something. The more specific your visual language, the closer the output lands to the trend you actually want.
Problem 4: The background looks AI-generated even if the face is okay
Add a real location reference. Instead of "urban background", write "Old Delhi street with colorful building facades and golden hour haze." Specificity forces the model out of its generic training average.
The 3-word fix for most bad outputs:
Add "ultra-realistic, photorealistic, 4K" to the end of any portrait prompt. It is not magic but it shifts the model's target significantly.
How to Create the Childhood AI Photo (Step by Step)
This trend is specifically about upload-based generation. Here is the exact process.
Open ChatGPT (free tier works, paid gives better results)
Start a new conversation
Click the paperclip icon and upload your childhood photo
In the same message, paste Prompt 1 or 2 from earlier in this article
If you want both childhood and current self in one image, upload both photos together with the prompt
Generate and download the result
If the face is wrong, add: "Do not change my facial structure. Keep my features exactly as they appear in the uploaded photo."
Gemini follows the same process. Go to gemini.google.com, upload your photo, paste the prompt.
Free vs Paid: What You Actually Need
Free options that work:
ChatGPT free tier: limited generations per day, image quality is good enough for most use cases
Gemini: currently free with a Google account, no daily cap that most users hit
Adobe Firefly: 25 free generative credits per month
Leonardo AI: 150 free tokens daily, resets every 24 hours
When to go paid: If you are generating for content creation professionally, the paid tiers of ChatGPT (Pro) and Midjourney (Basic at $10/month) give you enough volume and quality to produce at scale.
For most Instagram creators just doing it for fun or occasional posts, the free tiers are more than enough.
FAQ
Can I use these prompts for free?
Yes. ChatGPT free, Gemini free, Leonardo AI free tier, and Adobe Firefly free credits all work with these prompts. Paid tools give higher quality and more generations per day but are not required.
Why does my face look distorted in AI photos?
The model does not have a clear reference. Upload your photo directly in the same message as the prompt. Add the line "keep my face completely recognisable, no distortion, no AI-smoothing" to any portrait prompt.
Which is better, ChatGPT or Gemini?
For identity accuracy and keeping your face consistent across a collage: ChatGPT. For Indian aesthetics and cultural context: Gemini often reads it better. For pure image quality and artistic style: Midjourney.
How do I make the childhood AI photo look real?
Upload both photos (childhood and current) to ChatGPT in one message. Specify "photorealistic, keep original features, professional photography style, cinematic warm lighting." The more reference you give the model, the more accurate the result.
What aspect ratio should I use for Instagram Reels covers?
4:5 for feed posts and Reels covers. 9:16 for full-screen Stories. Add your preferred ratio at the end of any prompt: "Photo size 4:5" or "vertical format 9:16".
Can I do the Spiderman AI edit for free?
Yes. Use Prompt 11 or 12 from this article in ChatGPT free with a selfie uploaded. The result will not be as detailed as a Midjourney render but it is close enough for Instagram.
Why does my AI photo look like a painting instead of a photo?
You are missing realism anchors. Add "photorealistic, DSLR quality, natural lighting, shot on camera" to pull the output toward photography rather than illustration.
Want 500+ More Styles and Prompts?
These 50 prompts cover the trending styles right now. But there is a lot more to AI image generation than Instagram trends.
If you want to explore cinematic film styles, fine art aesthetics, fantasy and mythology, 3D and CGI looks, architectural photography, food and fashion, wildlife and nature -- all organised and ready to use across every major AI tool -- the guide I put together covers all of it.
500+ curated styles. Every prompt copy-paste ready. Tested across Midjourney, DALL-E, Stable Diffusion, Leonardo and more.
Get the AI Image Generation Guide here: aiunfiltered.gumroad.com/l/AIImageGenerationGuide
