Skip to content
PromptifyLab

Creators & Social

Viral AI Photo Trends prompts

7 prompts Free · no sign-up Works in ChatGPT, Claude & Gemini Search & filter these

Figurines, retro portraits, action-figure packs, photo restoration, pet art and miniature worlds. Below are 7 copy-ready prompts. Fill in the [BRACKETS], copy, and paste into ChatGPT, Claude, Gemini or any capable assistant.

Every few weeks a new AI photo style sweeps Instagram, TikTok and WhatsApp – collectible figurines, retro film portraits, boxed action figures. The prompts here are copy-ready versions of the formats that keep coming back, plus a prompt that reverse-engineers any new trend so you can try it on day one.

The 7 prompts

Beginner 2 blanks to fill

Turn a photo into a 3D collectible figurine scene

Create the popular figurine-on-a-desk style image from your own photo.

Prompt
(Attach a photo of yourself, your pet or your own product. Use an image tool that accepts reference photos, such as the image features in Gemini or ChatGPT.)

Create a 1/7 scale commercialised collectible figure of the subject in the attached photo, in a realistic style, placed on a computer desk in a real environment.

Details:
- The figure stands on a round, clear acrylic base with no text.
- Next to the desk, a toy packaging box printed with an original illustrated design of the same figure. Invent a name for the figure: [FIGURE NAME].
- On the computer screen behind it, show the 3D modelling process of the same figure.
- Lighting: soft indoor desk lighting, shallow depth of field, photographic realism.
- Keep the subject's key features, outfit colours and pose recognisable from the photo.
- Style details to add: [e.g. anime-inspired / realistic / chibi / superhero outfit]

Do not add any brand logos or trademarked characters.

What you get: A realistic photo-style image of a collectible figure of your subject on a desk, with packaging and a modelling screen.

Tip: Only use photos of yourself or people who have agreed. Keep brand logos and famous characters out - they are the usual reason these images get blocked or become unusable.

Open in Written for Gemini, ChatGPT · Reviewed September 25, 2026
Beginner 4 blanks to fill

Create a retro film-style portrait from a selfie

Restyle a portrait into a vintage 1990s film look with grain and warm tones.

Prompt
(Attach a clear, well-lit selfie or portrait of yourself.)

Transform the attached portrait into a retro 1990s film photograph.

Keep: the person's face, identity, hairstyle and expression exactly the same.
Change:
- Outfit: [e.g. flowing chiffon saree in deep red / oversized denim jacket / vintage suit]
- Setting: [e.g. warm-toned wall with a sunset glow / old city street / studio backdrop]
- Lighting: warm golden light from one side, soft shadows, slight glow
- Film look: visible grain, slightly faded colours, warm cast, gentle vignette
- Pose: [natural / looking over shoulder / holding a flower]
- Aspect ratio: [4:5 portrait]

Photorealistic, not a painting. No text, no watermark.

What you get: A vintage film-style portrait that keeps your face and changes the outfit, setting and lighting.

Tip: Say 'keep the face exactly the same' first. If the likeness drifts, re-run with a clearer, front-facing photo rather than adding more words.

Open in Written for Gemini, ChatGPT · Reviewed September 25, 2026
Beginner 5 blanks to fill

Make an action-figure blister pack of yourself

Create a toy-packaging image with accessories that describe your job or hobbies.

Prompt
(Attach a full-body or half-body photo of yourself.)

Create an image of an action figure of the person in the photo, sealed in a toy blister pack, like a store-bought collectible.

- The figure: same face, hair and outfit style as the photo, in a slightly toy-like plastic finish
- Pack header text: "[YOUR NAME]" and underneath "[YOUR ROLE, e.g. Digital Marketer]"
- Accessories in separate plastic slots: [LIST 3-5 ITEMS, e.g. laptop, coffee mug, camera, notebook]
- Packaging colours: [COLOURS]
- Background: plain studio background, product-shot lighting
- Aspect ratio: [3:4]

Check that all text is spelled exactly as given. No brand logos.

What you get: A product-shot style image of you as a boxed action figure with accessories and your name and role.

Tip: Text in images is still the weak spot. Keep the header to a few words, and if it misspells, regenerate or fix it in an editor.

Open in Written for Gemini, ChatGPT · Reviewed September 25, 2026
Beginner 2 blanks to fill

Restore and colourise an old family photo

Repair scratches and fading and add natural colour to a vintage photo.

Prompt
(Attach a scan or clear phone photo of the old picture. Photograph it flat, without glare.)

Restore this old photograph:
1. Remove scratches, dust, creases and stains.
2. Fix fading and improve contrast while keeping it natural.
3. Keep every face, expression and feature exactly as in the original - do not beautify, smooth skin or change ages.
4. [OPTIONAL] Colourise with realistic, period-appropriate colours. Known details: [e.g. her dress was dark green, the wall was whitewashed].
5. Keep the original framing. Do not add or remove people or objects.

Output a high-resolution version.

What you get: A cleaned and optionally colourised version of an old photo that keeps faces unchanged.

Tip: Tell it what colours you know. AI colourisation guesses, and family members will notice a wrong uniform or dress colour immediately. Always keep the original scan.

Open in Written for Gemini, ChatGPT · Reviewed September 25, 2026
Beginner 4 blanks to fill

Turn your pet into an illustrated character

Create cartoon, anime or storybook-style images of your pet.

Prompt
(Attach a clear photo of your pet.)

Create an illustration of the pet in the photo as a [STYLE: e.g. soft watercolour storybook / Pixar-like 3D cartoon / cosy anime / pixel art] character.

- Keep: fur colours, markings, eye colour and breed features recognisable
- Scene: [e.g. reading a book by a window / wearing a tiny astronaut suit on the moon / sleeping on a cloud]
- Mood: [cosy / heroic / funny]
- Aspect ratio: [1:1 for a profile picture / 4:5 for a post / 9:16 for a phone wallpaper]
- No text.

Make it charming and clean, suitable for printing.

What you get: A stylised illustration of your pet in a scene of your choice, keeping its markings.

Tip: Mention distinctive markings explicitly (a white patch on the left ear, one blue eye). Those are what make it look like YOUR pet.

Open in Written for Gemini, ChatGPT, Midjourney · Reviewed September 25, 2026
Beginner 4 blanks to fill

Create a miniature diorama of a place you love

Make a tilt-shift miniature world of a city, home or landmark.

Prompt
Create a detailed miniature diorama of [PLACE: e.g. my hometown's old market / a Karachi street at night / a cosy mountain cabin].

- Viewed from a 45-degree angle above, like a tabletop model
- Tiny people, vehicles and details: [LIST DETAILS THAT MAKE IT RECOGNISABLE]
- Tilt-shift photography look with shallow depth of field
- Time of day: [golden hour / night with warm lights / rainy morning]
- Presented on a wooden base on a table
- Aspect ratio: [4:5]

Photorealistic model-making style. No text or logos.

What you get: A tilt-shift miniature model of a place, with recognisable local details.

Tip: Local details are what make people share it: a specific shop sign style, a street food cart, a type of rickshaw. List three or four.

Open in Written for Gemini, ChatGPT, Midjourney · Reviewed September 25, 2026
Intermediate 1 blank to fill

Reverse-engineer any viral image style into a reusable prompt

Describe a viral image style so you can recreate it with your own subject.

Prompt
(Attach an example image of the style you want, or describe it in detail.)

Analyse this image's style and write a reusable prompt template for it.

Break down:
1. Subject treatment (realistic, stylised, toy-like, painted)
2. Composition and camera angle
3. Lighting (direction, softness, colour)
4. Colour palette
5. Texture and finish (grain, gloss, matte, film)
6. Background and props
7. Anything that makes it instantly recognisable

Then write:
- A PROMPT TEMPLATE with [SUBJECT] and other variables in brackets
- 3 filled-in examples using different subjects
- What NOT to copy: any logos, trademarks, a specific artist's signature style, or real people shown in the example

Keep the template original - capture the look, not the specific artwork.

What you get: A style breakdown and reusable prompt template with three filled examples and a list of things not to copy.

Tip: Trends move fast. Learning to break a style into lighting, composition and finish lets you recreate the next trend on day one.

Open in Written for Gemini, ChatGPT, Claude · Reviewed September 25, 2026

Where AI actually helps here

  • Restyling your own photo while keeping your face recognisable
  • Restoring and colourising old family photos
  • Turning pets and places into illustrated or miniature scenes
  • Breaking any new trend down into a reusable template

Where it falls down

  • Perfect likeness every time – some runs drift
  • Spelling text inside images reliably
  • Using other people’s photos without permission, or copying trademarked characters

The mistake almost everyone makes: Uploading photos of other people

Only use photos of yourself, or of people who have clearly agreed – and be careful with photos of children. Uploaded images may be processed and stored under the tool’s terms, so check its privacy settings first.

Free tool: Image Prompt Builder

Runs in your browser. No sign-up, nothing uploaded.

Open the Image Prompt Builder →

Questions people ask


Which AI app do these viral photo prompts work in?

Most need an image tool that accepts a reference photo, such as the image features in Gemini or ChatGPT. Results vary between tools and over time as models update.


Why doesn't the result look like me?

Use a clear, well-lit, front-facing photo, and say ‘keep the face exactly the same’ at the start of the prompt. Re-run a few times – results vary.


Is it safe to upload my photo to AI apps?

Check the app’s privacy settings and terms for how uploads are stored and used. Avoid uploading photos of children or sensitive images.