New Year photo prompts are a kind of photo-to-image AI prompt that turns your own uploaded picture into a New Year’s Eve portrait, a party scene or a “Happy New Year 2027” card in Google’s Gemini or OpenAI’s ChatGPT.
Each prompt below swaps the scene and the outfit but keeps a sentence that tells the tool to keep every uploaded face, and only the card prompts ask for readable words. Gemini and ChatGPT both take a photo on their free tiers, with usage limits on each.
Two Gemini rules decide whether a prompt works: editing an upload needs a personal account aged 18 or over, and merging several photos needs the model set to Flash or Pro.
The 46 prompts cover couples, families and kids, friends, solo portraits and profile pictures, fireworks, glam looks, jokes and cards, with saree and kurta looks across the groups. Skip them for any photo that must stay an exact record of the person.
Gemini stamps a visible watermark on its images and ChatGPT writes C2PA metadata into the file, so check the mark before you print a card. Find your group in the table, or read the three steps first.
| Who is in the picture | Prompts | Photos to upload | Setting to check in Gemini |
|---|---|---|---|
| Couples | 1-6 | One shared photo; two selfies for prompt 6 | 18 or over; Flash or Pro for two photos |
| Family and kids | 7-13 | One group photo, run from a parent’s account | 18 or over; editing is not available under 18 |
| Friends | 14-19 | One group photo; one per friend for prompt 19 | 18 or over; Flash or Pro for several photos |
| Solo and profile pictures | 20-25 | One photo | 18 or over; not Flash-Lite for repeated edits |
| Fireworks and countdown | 26-30 | One photo of one to four people | 18 or over |
| Glam and outfits | 31-36 | One photo | 18 or over; not Flash-Lite for repeated edits |
| Funny | 37-40 | One photo; a pet photo too for prompt 39 | 18 or over; Flash or Pro for two photos |
| Cards | 41-46 | One photo, or several for prompts 44 and 45 | 18 or over; Flash or Pro for several photos; Redo with Pro on a paid plan for lettering |
Sources: Gemini Apps Help on generating and editing images, checked November 2026.
In This Article (15 sections)
How to Use These New Year Photo Prompts
Every prompt here runs in three steps: upload one clear photo, paste the full prompt, then fix one detail per follow-up message.
- Upload a clear, front-lit photo. In Gemini, pick Create images from the tools menu, or on a computer open the sidebar and click Images, then add your picture. In ChatGPT, tap the + icon and select Add photos & files, or drag or paste the image in; it accepts PNG, JPEG and non-animated GIF files of up to 20MB each.
- Paste the whole prompt. Swap the bracketed parts first, such as [four] for your headcount, and leave the sentence about faces in place.
- Fix one detail per follow-up. Gemini takes requests to change the background, replace an object or add an element, and ChatGPT refines an image through ordinary conversation. Keep Gemini on Flash or Pro for those follow-ups, because Google says Flash-Lite isn’t designed for multiple edits.
The example images under the prompts are AI-generated with fictional people, made with OpenAI’s image generation tool, so a result built from a real upload will differ.
If you make solo images often, a Gemini avatar lets you create pictures of yourself without uploading a photo each time, and you can share it with friends or family.
The same three steps work on my Gemini photo prompts to copy for the rest of the year. The hardest New Year frame comes first: two people, often from two different phones, at midnight.
Couples New Year’s Eve Photo Prompts (1-6)
Prompts 1 to 5 start from one photo of the two of you, and prompt 6 builds the scene from two separate selfies, which in Gemini needs the model on Flash or Pro. Prompt 3 gives the rooftop party a saree and kurta.
For couple scenes outside the holiday, my Gemini couple photo prompts use the same face-keeping approach.
1. Midnight rooftop toast
Two glasses up on a rooftop as the sky lights.
Start in: Gemini or ChatGPT.
Create a photorealistic New Year's Eve photo from the uploaded picture of two people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly two people. They raise [champagne] glasses on a city rooftop at midnight, in a [dark suit] and a [red dress], fireworks bursting behind. Warm rim light on both faces. No text. 4:5 aspect ratio.

2. Black-tie gala staircase
The formal look, full length, with nothing to spell.
Start in: ChatGPT or Gemini.
Create an elegant black-tie portrait from the uploaded picture of two people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly two people. They pose on a curved ballroom staircase in a [tuxedo] and a [satin gown], gold balloons drifting overhead with no numbers on them. Soft chandelier light, full-length framing. 2:3 aspect ratio.

3. Saree and kurta rooftop party
A South Asian party look under string lights.
Start in: Gemini or ChatGPT.
Create a photorealistic rooftop party photo from the uploaded picture of two people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly two people. One wears a [navy sequined saree], the other an [ivory kurta with a Nehru jacket]. String lights overhead, fireworks above the skyline behind them. Warm evening glow. No text. 4:5 aspect ratio.

4. Pajama countdown at the window
For couples who stay in and watch from the sofa.
Start in: ChatGPT or Gemini.
Create a cozy photorealistic photo from the uploaded picture of two people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly two people. They sit on a sofa under one blanket in matching [striped] pajamas, watching fireworks through a big window. Lamp light inside, colored glow on their faces. No text. 3:2 aspect ratio.

5. Beach sparklers at midnight
Gold sparkler trails at the water’s edge.
Start in: Gemini or ChatGPT.
Create a photorealistic night beach photo from the uploaded picture of two adults. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly two people. They hold lit sparklers that leave gold trails, barefoot at the water's edge, [linen] outfits moving in the breeze. Sparkler light on their faces, dark sea behind. 3:2 aspect ratio.

6. Two selfies, one balcony
For a couple apart on the night, merged into one frame.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT; upload both selfies.
Create one photorealistic photo that combines the two uploaded selfies into a single scene with two people. Keep both faces, skin tones and hairstyles exactly as in their own photos, with exactly two people. They stand side by side on a balcony at midnight, city lights and distant fireworks behind. Matching light on both faces. No text. 4:5 aspect ratio.

A family photo adds more faces and a second question: whose account runs a child’s picture?
Family and Kids New Year Photo Prompts (7-13)
These seven prompts each state the headcount, and the four with children (9, 10, 11 and 13) should run from a parent’s adult account, because Google makes image editing unavailable to users under 18.

Children here hold sparkling juice or nothing, watch fireworks from behind glass and never hold a sparkler. Prompt 7 dresses three generations in kurtas and salwar kameez.
7. Three generations on the terrace
Grandparents in front, the whole family under fairy lights.
Start in: Gemini or ChatGPT.
Create a photorealistic family photo from the uploaded picture of [five] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [five] people. Grandparents sit in front, the others stand behind on a home terrace under fairy lights, in [kurtas and salwar kameez]. Glasses of sparkling juice in hand. Warm evening light. 3:2 aspect ratio.

8. Living-room countdown
Party hats and noisemakers, ready for the last ten seconds.
Start in: ChatGPT or Gemini.
Create a photorealistic living-room photo from the uploaded picture of [four] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [four] people. They wear paper party hats and blow noisemakers beside a wall clock with both hands near twelve and no numbers on the dial. Bright, playful indoor light. 3:2 aspect ratio.

9. Asleep before midnight
The family that never made it to twelve.
Start in: Gemini or ChatGPT, from a parent’s adult account.
Create a gentle photorealistic photo from the uploaded picture of [three] people, one adult and [two] children. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [three] people. All three are asleep on a sofa under a blanket, paper party hats tilted. Soft lamp light, a quiet room. No text. 4:5 aspect ratio.

10. Fireworks through the window
A child watching the show from a safe, warm room.
Start in: Gemini or ChatGPT, from a parent’s adult account.
Create a photorealistic photo from the uploaded picture of one child. Keep the face, skin tone and hairstyle exactly as in the photo, with exactly one person. The child, in [warm pajamas], watches fireworks from indoors with both hands on a closed window, colored reflections on the glass. Calm room behind. No text. 4:5 aspect ratio.

11. Noon balloon drop
A daytime countdown for small kids who go to bed early.
Start in: ChatGPT or Gemini, from a parent’s adult account.
Create a bright photorealistic playroom photo from the uploaded picture of [two] small children. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [two] people. Soft balloons in [gold and silver] float down around them as they laugh and reach up, fully dressed in [party outfits]. Daylight from a window. No text. 1:1 aspect ratio.

12. New Year’s Day walk
The morning after, in coats and low sun.
Start in: Gemini or ChatGPT.
Create a photorealistic morning photo from the uploaded picture of [three] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [three] people. They walk along a park path on New Year's Day in winter coats and scarves, frost on the grass, low sun behind the trees. Relaxed and candid. 3:2 aspect ratio.

13. Baby’s first New Year
A calm, safe resting scene in gold.
Start in: Gemini or ChatGPT, from a parent’s adult account.
Create a soft photorealistic portrait from the uploaded photo of one baby. Keep the face, skin tone and hair exactly as in the photo, with exactly one person. The baby lies awake on a flat padded mat with a [cream] blanket underneath, in a [gold knit outfit], a few soft balloons nearby. Gentle window light. No text. 4:5 aspect ratio.

Friends bring more faces, and often more photos to merge.
Friends and Party Photo Prompts (14-19)
Five of these six prompts start from one group photo; prompt 19 merges one photo per friend, so in Gemini it needs Flash or Pro. Ask everyone in the picture before you upload it: Google tells users to get permission before using other people’s faces in Gemini.
14. Confetti mid-air house party
The loudest second of the night, frozen.
Start in: Gemini or ChatGPT.
Create a photorealistic house party photo from the uploaded picture of [five] friends. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [five] people. Gold confetti bursts mid-air as they cheer in a living room, in [party outfits]. Flash-lit and energetic, with motion blur on the confetti only. No text. 3:2 aspect ratio.

15. Indo-western terrace party
A dance pose that mixes Indian and Western party wear.
Start in: ChatGPT or Gemini.
Create a photorealistic terrace party photo from the uploaded picture of [four] friends. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [four] people. They strike a dance pose under string lights in Indo-western outfits, such as a [cape-style kurta] and a [sequin crop top with a lehenga skirt]. Warm night light. No text. 3:2 aspect ratio.

16. Photo-booth strip
Four frames, four props, the same faces each time.
Start in: Gemini or ChatGPT.
Create a vertical photo-booth strip of four stacked frames from the uploaded picture of [three] friends. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [three] people in every frame. They change props each frame: glasses, a feather boa, a party horn, a group hug. Props carry no words or numbers. 2:3 aspect ratio.

17. Cabin deck at midnight
Snow, hot drinks and string lights on a railing.
Start in: ChatGPT or Gemini.
Create a photorealistic winter photo from the uploaded picture of [four] friends. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [four] people. They stand on a snowy cabin deck at midnight holding hot drinks, in knit hats and puffer jackets, string lights along the railing. Cold blue night, warm light on faces. 3:2 aspect ratio.

18. Street countdown crowd
Your group sharp in front, the city crowd soft behind.
Start in: Gemini or ChatGPT.
Create a photorealistic street photo from the uploaded picture of [four] friends. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [four] people in sharp focus at the front. Behind them a city street crowd is blurred, fireworks above the buildings. Arms up, cheering. Street light on their faces. No signs or text. 16:9 aspect ratio.

19. Reunion from separate photos
For friends in different cities who want one table.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT; upload one photo per friend.
Create one photorealistic dinner photo that combines the [four] uploaded photos into one scene with [four] people. Keep each face, skin tone and hairstyle exactly as in its own photo, with exactly [four] people. They sit around one candlelit table at a New Year's dinner, glasses raised, in [smart evening clothes]. Matching warm light on every face. 3:2 aspect ratio.
Some readers only want themselves in the frame, often for a new profile picture.
Solo New Year Portrait and Profile Picture Prompts (20-25)
Each of these six prompts needs one photo: square for a profile picture, taller for a post or a story. Prompts 21 and 22 are a women’s and a men’s party look; swap the bracketed outfit to change either.
20. Gold bokeh profile picture
A centered head-and-shoulders shot that survives a round crop.
Start in: Gemini or ChatGPT.
Create a photorealistic head-and-shoulders portrait from the uploaded photo of one person. Keep the face, skin tone and hairstyle exactly as in the photo. Set the person against soft gold bokeh lights, in a [black top], facing the camera with a relaxed smile. Even, flattering light, centered for a round profile crop. No text. 1:1 aspect ratio.

21. Sequin dress party portrait
A waist-up party shot in silver.
Start in: ChatGPT or Gemini.
Create a photorealistic waist-up party portrait from the uploaded photo of one woman. Keep her face, skin tone and hairstyle exactly as in the photo. Dress her in a [silver sequin dress] at a New Year's Eve party, blurred lights and guests behind. Soft key light on her face, sparkle on the fabric. No text. 4:5 aspect ratio.

22. Velvet blazer portrait
A dressed-up men’s look in a dim lounge.
Start in: Gemini or ChatGPT.
Create a photorealistic waist-up portrait from the uploaded photo of one man. Keep his face, skin tone, hairstyle and any beard exactly as in the photo. Dress him in a [burgundy velvet blazer] over a black shirt in a dim lounge, a glass in hand. Warm side light, shallow depth of field. No text. 4:5 aspect ratio.

23. Sparkler story portrait
A full-figure night shot sized for a story.
Start in: ChatGPT or Gemini.
Create a photorealistic night portrait from the uploaded photo of one adult. Keep the face, skin tone and hairstyle exactly as in the photo. The person holds a lit sparkler at arm's length on a quiet street, gold sparks lighting the face, in a [long coat]. Full figure, tall framing for a story. No text. 9:16 aspect ratio.

24. Window-seat fireworks
A quieter, reflective midnight.
Start in: Gemini or ChatGPT.
Create a calm photorealistic portrait from the uploaded photo of one person. Keep the face, skin tone and hairstyle exactly as in the photo. The person sits on an apartment window seat in [a cozy sweater], turned toward the glass as city fireworks bloom outside, colored light across the face. Dark room, reflective mood. No text. 4:5 aspect ratio.

25. Retro flash party snapshot
The on-camera flash look of an old party print.
Start in: ChatGPT or Gemini.
Create a retro party snapshot from the uploaded photo of one person. Keep the face, skin tone and hairstyle exactly as in the photo. The person laughs mid-moment at a crowded New Year's party, lit by direct on-camera flash, with film grain and slightly warm colors like a [1990s] print. No text. 3:2 aspect ratio.

Fireworks change the light on every face, so they get their own group.
Fireworks and Countdown Photo Prompts (26-30)
These five prompts ask the model to light the face separately from the sky. Each works from one photo of one to four people; change the bracketed headcount to match.
26. Waterfront skyline fireworks
A wide harbor shot with the skyline behind.
Start in: Gemini or ChatGPT.
Create a photorealistic waterfront photo from the uploaded picture of [two] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [two] people. Frame them waist-up at a harbor railing, a skyline behind and fireworks over the water. Light their faces with a soft separate key light so the sky does not wash them out. No text. 16:9 aspect ratio.

27. Clock tower at midnight
The countdown told by clock hands instead of digits.
Start in: ChatGPT or Gemini.
Create a photorealistic night photo from the uploaded picture of [two] friends. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [two] people. They look up at an old stone clock tower with both hands at twelve and no numerals on the dial, fireworks behind it. Lantern light on their faces. 2:3 aspect ratio.

28. Lake reflection fireworks
Fireworks twice: in the sky and in still water.
Start in: Gemini or ChatGPT.
Create a photorealistic lakeshore photo from the uploaded picture of [three] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [three] people. They stand on a wooden dock, fireworks above and mirrored in still water. A lantern at their feet lights their faces warmly. No text. 3:2 aspect ratio.

29. Confetti cannon at zero
The stage moment the count hits zero.
Start in: ChatGPT or Gemini.
Create a photorealistic celebration photo from the uploaded picture of [one] person. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [one] person. A stage confetti cannon fires behind at the stroke of midnight, streamers frozen mid-burst, arms raised. Bright stage light on the face. No text or numbers. 4:5 aspect ratio.

30. Long-exposure balcony streaks
Light trails behind a face that stays sharp.
Start in: Gemini or ChatGPT.
Create a photorealistic long-exposure night photo from the uploaded picture of one person. Keep the face, skin tone and hairstyle exactly as in the photo, sharp and still. The person leans on a balcony rail while firework light trails streak across the sky behind. A soft light holds the face steady. No text. 2:3 aspect ratio.

The outfit is the other half of a New Year look.
New Year Glam and Outfit Prompts (31-36)
These six prompts keep your face and pose and change only the clothes and backdrop, so you can preview a party look before you buy or borrow it. Prompt 32 is a sequined saree; prompt 36 swaps the outfit and nothing else.
Google’s overview says Gemini can take the texture, color or style of a reference photo and apply it to your subject. For festival sarees with diyas and rangoli, my Diwali photo prompts cover that look.
31. Gold sequin gown portrait
Party glamour next to a champagne tower.
Start in: Gemini or ChatGPT.
Create a glamorous photorealistic portrait from the uploaded photo of one woman. Keep her face, skin tone, hairstyle and pose as in the photo, and change only the clothes and setting. She wears a [gold sequin gown] beside a tower of champagne glasses at a party. Soft studio light with golden highlights. No text. 4:5 aspect ratio.
32. Sequined saree glam
Black and gold, lit like a studio shoot.
Start in: ChatGPT or Gemini.
Create a photorealistic studio portrait from the uploaded photo of one woman. Keep her face, skin tone, hairstyle and pose as in the photo, and change only the outfit and backdrop. She wears a [black-and-gold sequined saree] with a matching blouse and [small jhumka earrings], against a soft gold backdrop. Polished studio light. No text. 4:5 aspect ratio.
33. Classic tuxedo portrait
The men’s black-tie look on a dark backdrop.
Start in: Gemini or ChatGPT.
Create a photorealistic studio portrait from the uploaded photo of one man. Keep his face, skin tone, hairstyle and pose as in the photo, and change only the outfit and backdrop. He wears a [black tuxedo] with a [silver bow tie] against a dark studio backdrop. Low-key light with a soft rim on the shoulders. No text. 4:5 aspect ratio.

34. Disco-ball jumpsuit
Silver fabric under scattered spots of light.
Start in: ChatGPT or Gemini.
Create a photorealistic party portrait from the uploaded photo of one person. Keep the face, skin tone, hairstyle and pose as in the photo, and change only the outfit and backdrop. The person wears a [silver jumpsuit] under a disco ball that scatters small spots of light across the room. Fun, glossy color. No text. 4:5 aspect ratio.
35. Roaring Twenties look
Velvet, pearls and an art deco lounge.
Start in: Gemini or ChatGPT.
Create a 1920s-style party portrait from the uploaded photo of one person. Keep the face, skin tone, hairstyle and pose as in the photo, and change only the outfit and setting. The person wears [deep green velvet with pearls] in an art deco lounge with brass lamps. Warm, slightly soft vintage light. No text. 2:3 aspect ratio.

36. One-outfit swap
The same photo, background and pose, in a party outfit.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT.
Create an edited version of the uploaded photo of one person. Keep the face, skin tone, hairstyle, pose, background and framing exactly the same. Change only the outfit, to a [black velvet party dress] for New Year's Eve, with the fabric falling naturally. Do not change body shape, lighting or anything else in the photo. 4:5 aspect ratio.
Not every New Year image needs to look polished.
Funny and Resolution Photo Prompts (37-40)
These four joke prompts are for adults only. Leave children out of them, because OpenAI prohibits shaming or stigmatizing the body type or appearance of minors, and ask any adult you put in a joke first.
37. Asleep a minute before midnight
The party goes on without you.
Start in: Gemini or ChatGPT.
Create a funny photorealistic photo from the uploaded picture of one adult. Keep the face, skin tone and hairstyle exactly as in the photo. The person is fast asleep on a sofa in a crooked party hat, a noisemaker in hand, while a lively party carries on, slightly blurred, behind. Warm room light. No text. 3:2 aspect ratio.

38. Resolution day one
Last night’s outfit meets this morning’s treadmill.
Start in: ChatGPT or Gemini.
Create a funny photorealistic photo from the uploaded picture of one adult. Keep the face, skin tone and hairstyle exactly as in the photo. Still in last night's party outfit and a plain sash with no words on it, the person slumps over a treadmill handrail in a home gym, exhausted. Flat morning light. 4:5 aspect ratio.

39. Pet unimpressed by fireworks
You cheer, your pet does not.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT; upload your photo and your pet’s.
Create a funny photorealistic photo that combines the two uploaded photos: one adult and one [dog]. Keep the person's face, skin tone and hairstyle and the pet's markings exactly as in their photos. Both wear party hats by a window full of fireworks; the person cheers while the [dog] stares flatly at the camera. No text. 1:1 aspect ratio.

40. Diet vs dessert table
A small salad facing a very large cake.
Start in: Gemini or ChatGPT.
Create a funny photorealistic photo from the uploaded picture of one adult. Keep the face, skin tone and hairstyle exactly as in the photo. At a New Year's party buffet, the person holds a small salad while staring longingly at a huge dessert table of cakes and pastries. Bright party light. No signs or text. 4:5 aspect ratio.

A card adds words to the image, and words bring spelling checks.
Happy New Year 2027 Card Prompts (41-46)
All six card prompts ask for the exact text “Happy New Year 2027” (two add a family or team name) and no other words, so proofread every digit before you print or post.
On a paid Gemini plan, Redo with Pro regenerates the card with Nano Banana Pro, which Google says can add detail to images that use text. It stops working once you reach your daily quota of Nano Banana 2 images, so redo the card before the extras. Without a paid plan, keep the Gemini model setting on Flash, where Google says the Nano Banana 2 image model gives reliable text rendering.

Expect a mark on the result: Gemini uses an invisible SynthID watermark and a visible watermark, and OpenAI says all its generated images come with C2PA metadata, which sits in the file rather than on the picture.
41. Gold and black family card
Gold confetti on black, the family in the lower half.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT.
Create a New Year greeting card from the uploaded photo of [four] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [four] people. Place them in the lower half on a black background with falling gold confetti. Above them, add the exact text "Happy New Year 2027" in gold serif letters, and no other words. 5:7 aspect ratio.
42. Fireworks couple card
The greeting written across a night sky.
Start in: ChatGPT or Google’s Gemini on Flash or Pro.
Create a New Year greeting card from the uploaded photo of two people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly two people. They stand in front of a night skyline with fireworks. Across the sky, add the exact text "Happy New Year 2027" in white script, and no other words. 5:7 aspect ratio.

43. Minimal white name card
Lots of white space and the family name under the photo.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT.
Create a minimal New Year card from the uploaded photo of [three] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [three] people, shown small and centered on a white background. Below them, add the exact text "Happy New Year 2027 from the [Sharma] Family" in thin black letters, and no other words. 5:7 aspect ratio.

44. Photo collage card
Four photos kept apart in a grid, not merged.
Start in: Google’s Gemini on Flash or Pro, or ChatGPT, with all four photos attached.
Create a New Year collage card from the [four] uploaded photos, one photo per frame in a grid on a midnight blue background. Keep every face, skin tone and hairstyle exactly as in its photo, and do not merge or add people. Add the exact text "Happy New Year 2027" in gold across the center, and no other words. 7:5 aspect ratio.

45. Small team card
A card from a small business or team, with its name.
Start in: Google’s Gemini on Flash or Pro (needed if you upload one photo per person), or ChatGPT.
Create a New Year card for a small team from the uploaded photo of [six] coworkers. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [six] people. They stand together in an office hung with streamers. Above them, add the exact text "Happy New Year 2027 from [Acme Studio]" in white, and no other words. 7:5 aspect ratio.

46. Story-size greeting
A tall solo greeting for a story or status.
Start in: ChatGPT or Google’s Gemini on Flash or Pro.
Create a tall New Year greeting from the uploaded photo of one person. Keep the face, skin tone and hairstyle exactly as in the photo. The person stands in the lower third against a deep navy sky with soft gold sparkles. In the upper third, add the exact text "Happy New Year 2027" in gold letters, and no other words. 9:16 aspect ratio.

If your photo fits none of these, the prompts above share one structure you can reuse.
Build Your Own New Year Photo Prompt
Write six slots in this order: photo and headcount, identity sentence, scene, outfit, light and mood, then style and aspect ratio. The order follows Google’s formula for Gemini prompts (create, subject, action, scene, then composition, style, image quality and aspect ratio), so no prompt generator is needed.
| Slot | What to write | Example |
|---|---|---|
| Photo and headcount | How many people, from how many photos | Three friends, from one uploaded photo |
| Identity sentence | What must stay the same | Keep each face, skin tone and hairstyle exactly as in the photo |
| Scene | Where they are and what they do | Toasting on a houseboat deck as fireworks start |
| Outfit | Clothes, fabric, color | Navy and silver party clothes |
| Light and mood | One main light, one feeling | Lantern light on faces, excited |
| Style and ratio | The look, the framing, the shape | Google’s own example: in the style of an oil painting with 2:3 aspect ratio |
The blank template:
Create a [style] New Year image from the uploaded photo of [number] people. Keep each face, skin tone and hairstyle exactly as in the photo, with exactly [number] people. They are [scene]. They wear [outfit]. [Light and mood]. [Framing]. [ratio] aspect ratio.
Name the shape every time. Google’s example prompt asks for a 2:3 aspect ratio, and OpenAI lists aspect ratio among the specifics you can describe.
For jobs beyond the holiday, my AI image prompts by job use the same slots. A well-built prompt can still fail on the wrong tool setting.
Which Tool and Setting Fits Each Prompt Group
Pick by what your group needs: several photos, lettering or an under-18 user change the Gemini answer, while file size and image count matter in ChatGPT.
| Group | Gemini setting or gate | ChatGPT note | Watch for |
|---|---|---|---|
| Couples | 18 or over; Flash or Pro for two selfies | Add more images in later turns | Both faces drifting |
| Family and kids | Editing not available under 18 | Must not endanger anyone under 18 | Which account runs it |
| Friends | Flash or Pro to merge several photos | Image count per chat varies | A face missing |
| Solo and profile | Not Flash-Lite for repeated edits | Up to 20MB per image | Crop on a round avatar |
| Fireworks | 18 or over to edit an upload | Transforms uploaded images | Washed-out faces |
| Glam and outfits | Flash or Pro for repeated outfit edits | Refine through conversation | Pose or body changing |
| Funny | Flash or Pro for a person plus a pet | Heightened restrictions with real people in context | Consent for the joke |
| Cards | Redo with Pro on a paid plan; visible watermark | C2PA metadata on generated images | Every digit of 2027 |
Sources: Gemini Apps Help, Google’s Nano Banana overview, ChatGPT Image Inputs FAQ, OpenAI’s image generation announcement, checked November 2026.
Switch the Gemini model before you send the request. Nano Banana 2 runs on Flash or Pro and can accept multiple reference images. Nano Banana 2 Lite, on Flash-Lite, isn’t designed for multiple image references or multiple edits, which rules it out for prompts 6, 19, 39 and 44, and for prompt 45 when you upload one photo per person.

Nano Banana is available for free in Gemini, with limits, and a Google AI plan unlocks Nano Banana 2.1 and higher generation limits. The Gemini and ChatGPT pages I read do not state a free daily image count, so spend your first tries on the prompt you care about most.
My take
My pick by job: Gemini on Flash or Pro for any prompt that merges separate photos, because Google states that setting accepts multiple reference images, while OpenAI’s upload FAQ gives no fixed image count per conversation. For a card you will print or frame, I would lean to ChatGPT if Gemini’s visible watermark bothers you.
In either tool, run the prompt twice and compare both faces with your photo before you share one.
Past the free limits, my ranking of the best AI image generators by job compares paid tools. The right setting still sometimes returns a wrong face or a wrong year.
How to Fix a Changed Face, Wrong Headcount, Wrong Year or Soft Print
Find what went wrong in the table and send the instruction beside it in the same chat, so the tool keeps the image it already made.
| What went wrong | Send this, or do this |
|---|---|
| A face changed | “Keep everything, but match this face to the uploaded photo exactly.” In Gemini, switch from Flash-Lite to Flash or Pro |
| A person was added or lost | “There must be exactly [four] people. Restore the missing person and change nothing else.” |
| The year reads 2026 or is garbled | “Replace the text with exactly ‘Happy New Year 2027’ and keep the rest unchanged.” On a paid Gemini plan, try Redo with Pro |
| Hands or glasses look wrong | “Fix only the left hand holding the glass. Leave the rest of the image as it is.” |
| The image was refused or removed | Remove brand names and any real public figure, keep outfits modest, then rephrase the scene |
| The print looks soft | Measure the downloaded file with the method below |
Sources: Gemini Apps Help, checked November 2026.
Narrow fixes like these match a Nano Banana 2 feature Google calls local edits: quick, specific changes to part of an image. My ChatGPT photo editing prompts give more follow-up wording of this kind for OpenAI’s tool.
Both vendors state why a request can fail. Gemini Apps may remove images when Google’s systems detect a possible violation of its Terms of Service or Prohibited Use Policy, and OpenAI applies heightened restrictions when images of real people are in context.
Print sharpness depends on the downloaded file. Gemini’s help page says images download at 2K with a Google AI plan and at 1K without one, while Google’s Nano Banana overview lists 2K resolution under the free model. Neither page gives pixel counts, so measure your own file.
Take a hypothetical download that measures 1,024 pixels on its long edge, printed as a 4 x 6 inch card with a 6-inch long side: 1,024 / 6 = about 171 pixels per inch. A hypothetical 2,048-pixel file on the same card gives 2,048 / 6 = about 341 pixels per inch.
If your figure falls short of what your print shop asks for, keep the card for screens, or download again on a Google AI plan, which Gemini’s help page ties to 2K. A sharp, correct image still leaves the bigger question: whose photos to upload at all.
Is It Safe to Upload Family and Kids’ Photos?
It depends on your settings and on whose face is in the picture. With Gemini’s Keep Activity setting on, Google uses your activity to provide and improve its services, including training generative AI models, with the help of human reviewers.
Before an upload:
- Turn off Keep Activity, or use a temporary chat. Google keeps temporary chats, and chats made with Keep Activity off, for 72 hours to respond and to protect users, but temporary chats are not used to train Google’s AI models.
- Let an adult run it. On a personal account, Gemini needs 18 or over to generate and edit images and 13 or over to generate only, so a teen can make scenes from text but cannot edit a family photo.
- Ask every adult in the photo. Google asks users to get permission before using other people’s faces in Gemini, and OpenAI’s usage policies bar using someone’s likeness, including a photorealistic image, without consent in ways that could confuse authenticity.
- Keep children in ordinary scenes. OpenAI says its tools must not be used to endanger anyone under 18, which is why the kids’ prompts above keep children behind glass and away from sparklers.
- Say it’s AI when you share it. Google’s Generative AI Prohibited Use Policy bars misrepresenting generated content as made solely by a human in order to deceive.
- Read OpenAI’s data page before a ChatGPT upload. OpenAI’s image FAQ sends users to its page on how data is used to improve model performance, which lists the choices users have.
My take
For a child’s photo, I would upload from a parent’s account inside a Gemini temporary chat, since Google says those chats are not used to train its AI models. If another parent says no, run the prompt with the adults only. These are the vendors’ published rules, not legal advice.
New Year Photo Prompt FAQ
Are these prompts free to run?
Yes, within limits: Nano Banana is available for free in Gemini, where limits apply, and ChatGPT accepts image inputs on its Free plan, subject to usage limits.
Do they work without a photo?
Yes: delete the upload wording and the identity sentence, then describe the people. In Gemini, a personal account aged 13 or over can generate images this way, but the faces will be invented.
Can I change the year or the language on a card?
Yes: replace the quoted text in the prompt. Google lists better text rendering with more accurate spelling for all supported languages among Nano Banana 2’s features, but proofread every character, especially in a script you do not read.
Can I turn a New Year portrait into a video?
No, not with these prompts, which ask for still images. A moving clip needs motion and camera lines, and my Google Flow video prompts cover those.
Related Photo Prompt Pages
Two sibling pages use the same copy-and-paste format for other occasions:
- Christmas photo prompts: family portraits and holiday cards for the week before New Year’s Eve.
- Birthday photo prompts: a birthday portrait or card built around the day, for a January birthday right after the holiday.





