Build a Nano Banana (Gemini) or ChatGPT prompt that puts two people — a couple, friends, parent and child — into one realistic photo. Choose the pose, scene and light, copy the prompt, or blend it right here for free.
Say who the two people are. Everything is optional — the builder fills sensible defaults.
The pose is what makes two separate photos read as one moment.
Pick a scene or type your own. Light is what sells the merge.
Style, crop and outfits — leave the defaults for a normal photo.
Updates as you click. Copy it, or blend it on this page.
Free and unlimited. The prompt is built in your browser — nothing is uploaded until you choose to blend.
Every prompt the builder generates — whether you paste it into Nano Banana, Gemini, ChatGPT or use it here — follows the same structure. If you would rather write your own, this is the checklist:
That is the whole method. A prompt for merging two photos in Gemini needs nothing more exotic — no JSON, no “masterpiece, 8k” padding. Plain sentences in that order work best on every current model.
Each one is already in the 5-part format. Tap Use this prompt to load it into the builder’s output box, then copy it or blend it here. Swap the names, place or outfits to make it yours.
Using the two attached photos, create a realistic photograph of the two of us as a couple standing close together on a beach at sunset, my head resting on his shoulder, both smiling softly at the camera, warm golden-hour light, waist-up framing, 3:4 portrait. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Searchers ask for a “romantic couple prompt for Gemini AI” — this is that prompt, with the identity line added.Merge the people from these two photos into one realistic photo of a couple hugging warmly in a cosy café, soft window daylight, candid mid-laugh, keep the outfits from their own photos, waist-up framing. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create a realistic photograph of a mother from the first photo and her son from the second photo sitting together on a sofa in a bright living room, her arm around his shoulder, both relaxed and happy, soft natural daylight, full-body framing. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create one realistic photograph of the same person as an adult (first photo) and as a child (second photo) standing side by side and smiling at each other in a green park, soft daylight, waist-up framing. They are the same person at two ages. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
The viral “meet your younger self” trend — also on the Hug My Younger Self page.Create a realistic selfie-style photo of two close friends from these two photos, cheek to cheek and laughing, phone out of frame, on a rooftop with the city skyline behind, evening light, close-up framing, 1:1 square. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create a realistic wedding-album photograph of the bride from the first photo and her grandmother from the second photo standing together on a decorated wedding stage, the grandmother holding the bride’s hand, both smiling, warm indoor lighting, waist-up framing. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Using the two attached photos, create a Polaroid instant print with a white frame and slight camera flash of the two of us hugging in front of a plain light-grey backdrop, candid and playful, film grain, 1:1 square. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
The “Nano Banana Polaroid” prompt — there is a one-tap version on AI Polaroid Photo.Create a realistic photograph of a graduate in cap and gown from the first photo standing beside her father from the second photo at a graduation ceremony, his arm around her shoulder, both proud and happy, bright even daylight, full-body framing. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create a realistic photograph of two siblings from these photos standing together at a festival evening with strings of warm lights behind them, wearing traditional festive clothes, holding sparklers, happy and relaxed, waist-up framing, 3:4 portrait. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create a professional studio portrait of the two people from these photos standing side by side against a plain light-grey backdrop, matching smart formal outfits, bright even studio lighting, confident and calm, waist-up framing, 16:9 landscape. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create a black-and-white photograph of a grandfather from the first photo and his grandson from the second photo sitting together on a park bench, the grandfather’s hand on the boy’s shoulder, calm and warm, overcast even light, wide shot. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
Create a realistic, respectful photograph of the woman from the first photo standing beside her late mother from the second photo in a bright living room at home, close together with a gentle smile, soft natural daylight, waist-up framing, keeping the outfits from their own photos. Keep both faces exactly as in the source photos so each person is instantly recognisable, and match the light, shadows, skin tones and scale so it looks like one real photograph. No extra people, no text.
More on this on the Memorial Photo Merge page.The same prompt works in all three; the difference is how you attach the photos and what each one costs.
| Gemini app (Nano Banana) | ChatGPT | AIMerger (this page) | |
|---|---|---|---|
| How to add photos | Tap + → add both photos, then paste the prompt | Click the paperclip → upload both photos, paste the prompt | Drop both photos in the tool below — the prompt is already filled in |
| Account | Google account | OpenAI account | None |
| Free use | Daily limit on image generation | Small daily limit on the free plan | 1 free blend a day, then credits from $4.99 that never expire |
| Face accuracy | Very good when the prompt names each photo | Good, tends to soften faces | Same Nano Banana model as Gemini, tuned for merging people, up to 4K |
| Watermark | Small Gemini mark | None | None |
Set Paste it into in the builder to the app you will use: the opening sentence changes to “Using the two attached photos…” for Gemini and “Using the two uploaded photos…” for ChatGPT, which is how those apps expect to be told where the people come from.
Once the two-people prompt works, the same formula covers the other edits people search for. Each of these has its own one-tap page on AIMerger, so you don’t have to type anything:
Looking for a bigger library? The AI photo merge prompts page has 40 copy-paste prompts sorted by tool.
Ready? Scroll up, build the prompt, and press Blend it here — or paste it into Gemini or ChatGPT. For the step-by-step version with screenshots, read how to put two people in one photo.
Drop in 2–8 pictures — merged instantly on your device, never uploaded.
📸 Drop 2–8 photos here, or click to choose
JPG, PNG or WebP · merged instantly in your browser · nothing is uploaded anywhere
Side-by-side, grid or collage — tweak spacing, corners and size live.
Full quality, no watermark — the button turns green when ready.
One photo + the options below works great — add 2+ when you want to truly merge people or images.
✨ Drop 1–4 photos here, or click to choose
One photo: change outfits, hair, background, or see your future baby · Two+: merge people and scenes into one picture
One tap sets everything up — 23 styles, each with its own quick options.
Pick a scene, lighting and quality — or just describe what you imagine.
Standard 1K costs 1 credit. 2K costs 2 credits and 4K costs 4 — you’ll be asked to confirm before a bigger image is generated.
10–40 seconds. Failed merges never cost a try or a credit.
No key yet? Get a credit pack — your key is shown on screen the moment you pay (and emailed as a backup). Credits never expire.
The best prompts name five things: who the two people are, how they interact (side by side, hugging, holding hands), the scene, the light, and an instruction to keep both faces exactly as in the source photos. The builder above writes all five for you; a one-line prompt like “put us together” leaves the AI to guess the rest.
Yes. Choose “Gemini app (Nano Banana)” in the builder, attach both photos in Gemini with the + button, paste the prompt and send. Nano Banana keeps faces well when the prompt tells it which photo each person comes from, which the builder does.
Yes. Pick “ChatGPT” in the builder, upload both photos in the chat and paste the prompt. ChatGPT’s image model tends to soften faces, so keep the “faces exactly as in the source photos” line and ask again if a face drifts.
Yes — building and copying prompts is free and unlimited, and nothing is uploaded until you choose to blend. Blending on AIMerger includes one free AI blend a day with no signup; credit packs never expire.
Yes. Choose “a couple”, a pose such as “hugging warmly” or “standing close, side by side”, and a scene. Use two front-facing selfies taken in similar light; the guide to a couple photo from two selfies has more tips.
Usually because the prompt did not tell the model to keep the faces, or the source photos were small, blurry or heavily filtered. Every prompt from the builder ends with an identity line — keep it — and start from clear, well-lit, unfiltered photos of each person.
Everything on AIMerger brings things together — here is what pairs with this one.