Prompt Library

ChatGPT Image Prompts to Create Stunning AI Photos

48 copy-paste prompts

48 copy-paste ChatGPT prompts for AI selfies, product photography, photo editing, social media visuals, viral image styles, and creative artwork - no Photoshop required.

In short: This page contains 48 copy-paste ready prompts, organized into 7 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

Louis Corneloup
By Louis Corneloup · Founder, Techpresso
Last updated ·Hand-curated & tested by the AI Academy team

AI Selfies & Portraits

6 prompts

Professional Headshot from Description

1/48

✨ What it does

ChatGPT generates a LinkedIn-quality headshot of a [gender] with [hair color], [hair length] hair, [skin tone] skin, and [eye color] eyes, on a [solid neutral / blurred office / outdoor] background. You generate a few and pick the one that looks most like you.

Generate a professional headshot photo of me based on this description: I am a [gender] with [hair color], [hair length] hair, [skin tone] skin, and [eye color] eyes. I am [age range]. Style the photo as a LinkedIn-quality headshot with a [solid neutral / blurred office / outdoor] background. I am wearing [clothing description]. The lighting should be soft and flattering, shot at eye level with a slight smile. Make it look like it was taken with a DSLR at f/2.8 with natural bokeh.

💡

Pro tip: Include specific details about your jawline, face shape, and expression to get a more accurate result. Vague descriptions produce generic faces.

Consistent Character Across Multiple Scenes

2/48

✨ What it does

ChatGPT generates a photo of the same person, using your [detailed physical description including age, build, hair, skin, distinguishing features], in a new scene you describe. You generate a few scenes and pick the ones that still look like the same person.

I want to create a series of images of the same person in different settings. The character is: [detailed physical description including age, build, hair, skin, distinguishing features]. Generate this exact person in the following scene: [scene description, e.g., "sitting at a coffee shop with a laptop, morning light coming through the window"]. Maintain absolute consistency in their facial features, body proportions, and overall appearance. The style should be [photorealistic / illustrated / cinematic].

💡

Pro tip: Save your character description as a reusable block. Paste it into every prompt to maintain consistency across your entire image series.

Couples or Group Portrait

3/48

✨ What it does

ChatGPT generates a [couples portrait / family photo / group photo] of the people you describe, posed together in the setting you name. You generate a few and pick the grouping that looks natural.

Create a [couples portrait / family photo / group photo] with these people: Person 1: [detailed description] Person 2: [detailed description] [Add more if needed] Scene: [describe setting, e.g., "standing in front of autumn foliage in a park"]. They are [posing description: arms around each other, sitting on a bench, laughing together]. The mood is [warm and candid / formal and posed / playful]. Lighting: [golden hour / studio / natural overcast]. Style: [photorealistic portrait photography with shallow depth of field].

💡

Pro tip: Describe the relationship between people through their body language in the prompt - "leaning into each other" produces more natural results than just listing positions.

Seasonal or Holiday Self-Portrait

4/48

✨ What it does

ChatGPT generates a [holiday/season] portrait of the person you describe, in a setting like a tree, pumpkin patch, or snowy porch, wearing [seasonal clothing description]. You generate a few and pick one for the holiday post.

Generate a [holiday/season] themed portrait photo. The person is: [physical description]. The setting is [describe: in front of a Christmas tree, at a pumpkin patch, cherry blossoms in spring, snowy cabin porch]. They are wearing [seasonal clothing description]. Include [seasonal elements: string lights, falling leaves, snowflakes, flowers]. The color palette should emphasize [warm tones / cool tones / pastels]. Style: lifestyle photography with [cozy / festive / bright and airy] mood. Aspect ratio: [4:5 for Instagram / 16:9 for cover photo].

💡

Pro tip: Generate seasonal content 2-3 weeks before the holiday. Early holiday posts get more engagement than posts on the actual day when feeds are oversaturated.

Before and After Transformation Photo

5/48

✨ What it does

ChatGPT generates a side-by-side before-and-after of a [type of transformation: fitness, style makeover, office renovation, room redesign], with the same camera angle on both sides. You generate a few and pick the pair that makes the change obvious.

Create a side-by-side before-and-after image showing a [type of transformation: fitness, style makeover, office renovation, room redesign]. Left side (before): [describe the "before" state in detail]. Right side (after): [describe the "after" state in detail]. Both images should have the same camera angle, lighting conditions, and composition so the transformation is the focal point. Add a subtle dividing line between the two halves. Style: clean, realistic photography with even lighting.

💡

Pro tip: Keep the camera angle and framing identical between both halves. The power of before/after comes from the contrast, not the composition change.

Professional Bio Photo Set

6/48

✨ What it does

ChatGPT generates a set of three matching bio photos of [name/description]: a close headshot, a half-body shot, and a wider environmental frame. You generate a few sets and pick the trio that looks like the same person.

Generate a set of 3 professional bio photos for [name/description]. Physical appearance: [detailed description]. The three photos should be: 1. Close-up headshot: shoulders up, direct eye contact, confident slight smile, [solid/gradient] background 2. Half-body environmental shot: sitting or standing in a [relevant professional setting], looking slightly off-camera, natural expression 3. Action shot: actively doing [relevant professional activity], candid feel but well-composed All three should have consistent color grading ([warm / cool / neutral]), the same person with identical features, and a [modern and clean / warm and approachable / bold and creative] aesthetic.

💡

Pro tip: Use the close-up for LinkedIn, the environmental shot for your website bio, and the action shot for speaker profiles or press kits.

Photo Editing & Enhancement

6 prompts

Background Replacement

7/48

✨ What it does

ChatGPT generates [describe the person or object in detail, including clothing, pose, expression] on a new background after you remove [describe what you want removed], with matching light and shadow. You generate a few swaps and pick the one where the edges look natural.

Take this photo concept and generate it with a different background. Subject: [describe the person or object in detail, including clothing, pose, expression]. Original background: [describe what you want removed]. New background: [describe the replacement background in detail, e.g., "a modern office with floor-to-ceiling windows overlooking a city skyline"]. Match the lighting direction and color temperature of the new background to the subject. Ensure shadows fall naturally. The result should look like the subject was actually photographed in the new location.

💡

Pro tip: Specify the direction of light on both the subject and the new background. Mismatched lighting is the #1 giveaway of a bad composite.

Color Grading and Mood Shift

8/48

✨ What it does

ChatGPT generates two versions of [describe the photo scene in detail, including subjects, composition, and setting], one in [mood A] color grading and one in [mood B]. You compare the pair and pick the mood you want.

Generate two versions of this scene: [describe the photo scene in detail, including subjects, composition, and setting]. Version 1: [mood A] color grading - [describe: warm golden tones, high contrast, rich shadows, like a Wes Anderson film] Version 2: [mood B] color grading - [describe: cool blue tones, desaturated, moody shadows, like a David Fincher film] Keep the composition and content identical between versions. Only the color treatment, contrast, and mood should change. This will help me decide which aesthetic works best for my [project type].

💡

Pro tip: Choose your color grade before your photoshoot (or generation session), not after. It affects wardrobe, prop, and setting decisions.

Image Upscale and Detail Enhancement

9/48

✨ What it does

ChatGPT recreates your low-quality scene as a sharp [photorealistic / illustrated] image with clearer detail and texture. You generate a few and pick the one that matches the original composition.

I have a low-quality image concept I want to recreate at high resolution. The original image shows: [describe every detail of the image you want to recreate - subjects, composition, colors, lighting, textures, background elements]. Recreate this scene in [photorealistic / illustrated] style at the highest possible quality. Add realistic detail to: skin texture, fabric weave, environmental textures (wood grain, metal reflections, foliage). The lighting should be [describe]. Resolution: maximum quality, sharp focus throughout.

💡

Pro tip: List every texture you can see in the original: fabric type, surface materials, natural elements. More texture details produce more realistic outputs.

Object Removal and Scene Cleanup

10/48

✨ What it does

ChatGPT generates a clean version of [describe the scene] with [list objects to remove] gone and the gaps filled to match the surroundings. You generate a few and pick the one where the fill looks natural.

Generate a clean version of this scene: [describe the scene] but WITHOUT these elements: [list objects to remove]. The areas where removed objects were should be filled with [describe what should be there instead: continuation of the background, empty table surface, clean wall, etc.]. The fill should match the surrounding textures, lighting, and perspective perfectly. No visible seams or artifacts. Keep everything else in the scene exactly as described.

💡

Pro tip: Be specific about what should fill the empty space. "Remove the lamp" is less useful than "remove the lamp and extend the bookshelf behind it."

Lighting Correction

11/48

✨ What it does

ChatGPT regenerates [describe the scene in detail] with corrected lighting after you name the problem, such as harsh shadows or a backlit subject. You generate a few and pick the version that looks professionally lit.

Generate this scene with corrected lighting: [describe the scene in detail]. Current lighting problem: [describe: too dark, harsh shadows, flat lighting, backlit subject, mixed color temperatures]. Corrected version should have: [describe ideal lighting - e.g., "soft key light from the left at 45 degrees, gentle fill from the right, warm color temperature around 4500K"]. The subject should be properly exposed with detail visible in both highlights and shadows. Add [catchlights in eyes / rim light / hair light] if a person is in the scene.

💡

Pro tip: Learn the names of basic lighting setups (Rembrandt, butterfly, split, loop) so you can reference them in prompts for consistent professional results.

Style Transfer Between Photos

12/48

✨ What it does

ChatGPT generates the subject and scene you want to keep, in the visual style of a reference look you name, such as vintage film or a painterly grade. You generate a few and pick the one that keeps the scene but wears the style.

Generate an image that combines these two concepts: Content: [describe the subject and scene you want to keep - e.g., "a woman sitting at a cafe reading a book"] Style: Apply the visual style of [describe the reference aesthetic - e.g., "1970s Kodachrome film photography with warm vintage tones, film grain, slightly faded highlights, and rich earth tones"] The content and composition should match the first description exactly, but the color palette, texture, grain, contrast, and overall mood should match the style reference. The result should look like the content was photographed using the style reference's techniques and equipment.

💡

Pro tip: Reference specific film stocks (Portra 400, Kodachrome 64, Velvia 50) or photographer names for more precise style transfer.

Product Photography

6 prompts

Clean Product Shot on White Background

13/48

✨ What it does

ChatGPT generates a studio product photo of [product description: type, color, material, size, distinguishing features] on a pure white background, shot at a [slight angle / straight-on / 3/4 view]. You generate a few and pick the listing-ready shot.

Generate a professional product photo of [product description: type, color, material, size, distinguishing features]. Place it on a pure white (#FFFFFF) background. The product should be centered, photographed at a [slight angle / straight-on / 3/4 view] to show its best features. Lighting: soft studio lighting with no harsh shadows, a subtle shadow beneath the product for grounding. The image should look like an Amazon or Shopify product listing photo. Show the product at [actual size relative to frame / filling 80% of frame]. Focus should be tack-sharp on the product.

💡

Pro tip: Generate multiple angles (front, side, 3/4, top-down) of the same product in one session to build a complete listing gallery.

Lifestyle Product Photo

14/48

✨ What it does

ChatGPT generates a lifestyle photo of [product description] in use in a [setting: modern kitchen, cozy living room, outdoor patio, gym, office desk], with [description or "not visible - just hands"]. You generate a few and pick the candid one.

Create a lifestyle product photo showing [product description] being used naturally in [setting: modern kitchen, cozy living room, outdoor patio, gym, office desk]. The person using it is [description or "not visible - just hands"]. The scene should feel candid and aspirational, not staged. Include environmental details: [list 3-4 props that complement the product, e.g., "a cup of coffee, a plant, morning newspaper"]. Lighting: [natural window light / golden hour / soft ambient]. The product should be clearly visible but not awkwardly centered. Style: editorial lifestyle photography for a brand like [reference brand].

💡

Pro tip: The best lifestyle shots make the product a natural part of the scene, not the obvious center of attention. Let the environment sell the lifestyle.

Product Flat Lay Composition

15/48

✨ What it does

ChatGPT generates a top-down flat lay of [main product] with the props you list, arranged on a [surface: marble countertop, wooden table, linen fabric, colored paper]. You generate a few layouts and pick the one that tells the story.

Generate a top-down flat lay photo featuring [main product] surrounded by [list 5-8 complementary items that tell a story, e.g., for a journal: "fountain pen, coffee cup, reading glasses, dried flowers, wax seal, vintage stamps"]. Arrange them on a [surface: marble countertop, wooden table, linen fabric, colored paper]. The composition should follow [rule of thirds / symmetrical / organized chaos] layout. Color palette: [describe 3-4 dominant colors]. Lighting: even, soft overhead light with minimal shadows. Leave [some / no] negative space. Style: Instagram-worthy flat lay for a [brand type] aesthetic.

💡

Pro tip: Use odd numbers of objects (3, 5, 7) in flat lays. Odd groupings create more visual tension and feel less staged than even numbers.

Product Scale and Size Reference

16/48

✨ What it does

ChatGPT generates a product photo of [product description] next to a [hand, coin, smartphone, coffee mug, ruler] so the [measurements] are obvious. You generate a few and pick the one that answers the size question.

Generate a product photo of [product description] that clearly communicates its size. Show it: (1) next to a common size reference object: [hand, coin, smartphone, coffee mug, ruler], (2) in a setting where familiar objects provide scale context. The product dimensions are approximately [measurements]. Make sure the reference object is universally recognizable. The photo style should be clean and informative, suitable for an e-commerce listing where customers frequently ask "how big is it?" Lighting: bright, even, detail-revealing.

💡

Pro tip: A human hand holding the product is the most universally understood size reference and also makes the product feel tangible and real.

Product Packaging Mockup

17/48

✨ What it does

ChatGPT generates a packaging mockup for [product name], a [box type / bag / bottle / pouch / tube] with your colors, logo, and finish. You generate a few angles and pick one for the deck or pre-production review.

Generate a product packaging mockup for [product name]. The packaging is a [box type / bag / bottle / pouch / tube] with these design elements: [describe: colors, logo placement, text, patterns, material finish]. Show the packaging: (1) closed and standing upright at a 3/4 angle, (2) partially open revealing the product inside. The setting is [on a shelf / on a table / floating on a gradient background]. Include subtle reflections and realistic material textures ([matte / glossy / kraft paper / metallic]). The packaging should look premium and ready for a brand presentation.

💡

Pro tip: Show packaging in context (on a shelf next to competitors, in a customer's hands) to test how it stands out in the real world, not just in isolation.

Multi-Angle Product Gallery

18/48

✨ What it does

ChatGPT generates a [4 / 6]-image gallery of the same [product description] from the front, side, top, and a close-up of [key feature: texture, clasp, label, stitching]. You generate the set and pick the angles that match your listing.

Create a [4 / 6]-image product gallery for [product description]. Each image should show the same product from a different angle: 1. Front-facing hero shot at slight angle 2. Side profile view 3. Top-down view 4. Close-up detail shot highlighting [key feature: texture, clasp, label, stitching] [5. Back view if applicable] [6. In-use or in-context shot] All images should have: identical white background, consistent studio lighting, same color temperature, and the product at the same relative size in frame. This should look like a professional e-commerce product gallery.

💡

Pro tip: Consistent lighting across all angles is more important than perfect lighting on any single shot. Mismatched gallery photos look amateur.

Social Media Visuals

6 prompts

Instagram Post Visual

19/48

✨ What it does

ChatGPT generates a square or tall Instagram image about [topic] in your [describe: minimalist, bold, colorful, earthy, luxury] style, with space for text. You generate a few and pick the one that stops the scroll.

Generate an Instagram-optimized image (1080x1080 or 4:5 aspect ratio) for a post about [topic]. The visual style should match my brand: [describe: minimalist, bold, colorful, earthy, luxury]. The image should include: [describe scene or concept]. Color palette: [list 3-4 colors]. Text overlay space: leave a [top / center / bottom] area relatively clean for potential text overlay. The image should stop the scroll - it needs to be visually striking even at thumbnail size. Style reference: like what you would see on [@reference account] or [brand name].

💡

Pro tip: View every image at the size of a phone screen thumbnail before using it. If it does not pop at 150x150 pixels, it will not stop the scroll.

LinkedIn Banner Image

20/48

✨ What it does

ChatGPT generates a wide LinkedIn banner for a [job title] in [industry], visualizing [describe: innovation, leadership, data-driven decisions, creative problem solving]. You generate a few and pick the one that sits well next to your headshot.

Generate a LinkedIn banner image (1584x396 pixels, or 4:1 aspect ratio). I am a [job title] in [industry]. My personal brand is about [describe: innovation, leadership, data-driven decisions, creative problem solving]. The banner should include: [visual elements that represent your brand - e.g., "abstract data visualization patterns", "cityscape with modern architecture", "clean workspace with technology"]. Color palette: [professional colors that complement my headshot]. Leave the left 30% relatively empty as my profile photo overlaps there. Style: professional, modern, and distinctive without being flashy.

💡

Pro tip: Your LinkedIn banner is the largest visual on your profile. Make it communicate one clear message about what you do or believe, not a collage of everything.

YouTube Thumbnail

21/48

✨ What it does

ChatGPT generates a 16:9 YouTube thumbnail for "[video title]" about [topic], with a bold face, a key object, and [2-3 bold, contrasting colors]. You generate a few and pick the one that still reads when it is tiny.

Generate a YouTube thumbnail (1280x720, 16:9) for a video titled "[video title]" about [topic]. The thumbnail should include: [describe: a person with [expression] on the [left/right] side, a key visual element on the opposite side]. Use [2-3 bold, contrasting colors]. The emotional hook is [curiosity / shock / excitement / transformation]. Leave space for text overlay that will read: "[3-5 word text]". The image should be readable and impactful at 168x94 pixels (YouTube small thumbnail size). Style: high contrast, saturated, attention-grabbing. Background: [describe or "blurred/bokeh"].

💡

Pro tip: Squint at your thumbnail from across the room. If you can still understand the concept, it will work at small sizes. If you cannot, simplify it.

Story or Reel Cover Image

22/48

✨ What it does

ChatGPT generates a tall [9:16 aspect ratio] Story or Reel cover about [topic], with [what the main image shows] on a [solid color / gradient / blurred photo / pattern]. You generate a few and pick the cover that makes people tap.

Generate a [9:16 aspect ratio] image for an Instagram Story or Reel cover about [topic]. The design should be [describe aesthetic]. Central visual: [what the main image shows]. Background: [solid color / gradient / blurred photo / pattern]. The image should work as a static cover that makes people want to tap. Include visual space for a [title / question / CTA] text overlay. Brand colors: [list]. The composition should draw the eye to the center-top of the frame where Instagram shows the content. Style: [polished and professional / casual and authentic / bold and graphic].

💡

Pro tip: Reel covers are the new first impression on your profile grid. Design them as a cohesive series with a consistent template, not random screenshots.

Quote Card Visual

23/48

✨ What it does

ChatGPT generates a quote card for "[paste quote]" by [author/source], in a [1:1 / 4:5 / 16:9] frame with the background and [serif / sans-serif / handwritten] look you choose. You generate a few and pick the one you would actually share.

Generate a visually striking quote card image for this quote: "[paste quote]" attributed to [author/source]. Design specifications: aspect ratio [1:1 / 4:5 / 16:9], background [describe: textured paper, gradient, photo-based, abstract pattern], font style suggestion [serif / sans-serif / handwritten], color palette [describe]. The quote should be the focal point with clear visual hierarchy. Include the attribution in smaller text below. Add [subtle decorative elements: quotation marks, line dividers, small icons] without cluttering. The overall feel should be [inspirational / thought-provoking / bold / elegant].

💡

Pro tip: The best quote cards have 15 words or fewer. If your quote is longer, pull out the single most powerful sentence.

Carousel Slide Series

24/48

✨ What it does

ChatGPT generates a [5 / 7 / 10]-slide carousel about [topic], from a hook image through one idea per slide to a last-slide ask. You generate the series and pick the flow that stays on brand.

Generate [5 / 7 / 10] carousel slide images for a post about [topic]. Each slide should be 1080x1080 (1:1). The series needs: Slide 1 (hook): [bold visual that stops the scroll, teases the topic] Slide 2-[N-1] (content): [describe what each slide covers - one concept per slide] Final slide (CTA): [call to action - follow, save, share, link in bio] All slides must have: consistent brand colors [list], the same visual template/layout, clear readable text areas, and a cohesive flow where each slide makes you want to swipe to the next. Style: [educational / storytelling / data-driven / tips-based]. Include slide numbers or progress indicators.

💡

Pro tip: The swipe rate from slide 1 to slide 2 determines whether Instagram pushes your carousel. Make slide 1 a cliffhanger that demands the swipe.

Creative & Artistic

5 prompts

Specific Art Style Rendering

25/48

✨ What it does

ChatGPT generates [describe your scene or subject] in a named art style, with the details that make that movement recognizable. You generate a few and pick the one that looks like the real style, not a generic filter.

Generate an image of [describe your scene or subject] in the exact style of [art movement or artist style: Art Nouveau, Bauhaus, Studio Ghibli, Moebius, Art Deco, Impressionism, Cyberpunk, Ukiyo-e]. Include the defining characteristics of this style: [list 3-4 specific elements, e.g., for Art Nouveau: "organic flowing lines, floral motifs, muted gold and green palette, decorative borders"]. The composition should follow [describe layout]. Color palette: [authentic to the style]. The result should be indistinguishable from actual work created in this style.

💡

Pro tip: Name specific characteristics of the art style rather than just the style name. "Art Deco" is vague. "Art Deco with geometric gold patterns, stepped forms, and rich jewel tones" is specific.

Surreal Photo Manipulation Concept

26/48

✨ What it does

ChatGPT generates a photorealistic picture of an impossible scene you describe, lit as if it were a real photograph. You generate a few and pick the surreal image that still looks grounded.

Generate a surreal photograph that combines [impossible concept: e.g., "a staircase that loops back on itself like an Escher drawing", "a person holding the moon like a balloon", "a room where gravity works differently on each wall"]. The style should be photorealistic despite the impossible subject matter. Lighting should be natural and consistent as if this scene were actually photographed. Include small realistic details that ground the surreal elements (texture on surfaces, natural shadows, atmospheric haze). The mood is [dreamlike / unsettling / whimsical / contemplative]. Aspect ratio: [specify].

💡

Pro tip: The more realistic the non-surreal elements are, the more impactful the surreal element becomes. Contrast is what makes the impossible feel uncanny.

Digital Illustration in Specific Medium

27/48

✨ What it does

ChatGPT generates an illustration of [describe scene] that looks like watercolor, oil, pencil, ink, or another medium you name, including the messy texture. You generate a few and pick the one that feels handmade.

Create an illustration of [describe scene] that looks like it was created with [specific medium: watercolor on cold-pressed paper, oil paint on canvas, colored pencil on toned paper, ink wash, gouache, charcoal on newsprint]. Include the texture and imperfections of this medium: [e.g., for watercolor: "visible brush strokes, pigment granulation, soft edges where colors bleed, white paper showing through in highlights"]. The subject is [describe]. Color palette: [limited to what this medium naturally produces]. The illustration should look like a finished piece by a skilled artist, not a digital filter.

💡

Pro tip: Reference the limitations of the medium in your prompt. Watercolors can not produce pure black; charcoal creates a specific kind of softness. These limitations are what make it look real.

Comic or Graphic Novel Panel

28/48

✨ What it does

ChatGPT generates a [single panel / 4-panel strip / full page layout] comic of your characters in the scene you describe, in a style such as manga or a graphic novel. You generate a few and pick the panel that tells the beat.

Generate a [single panel / 4-panel strip / full page layout] comic illustration. Style: [describe: classic superhero, manga, indie graphic novel, Franco-Belgian, underground comix, web comic clean line]. Characters: [describe each character's appearance]. Scene: [describe what is happening]. Include [speech bubbles with text: "(dialogue here)" / no text / sound effects]. The art should include [hatching / screentones / flat colors / painted]. Panel borders: [thick black / thin / borderless / dynamic broken frames]. Aspect ratio and layout: [describe].

💡

Pro tip: Describe the "camera angle" for each panel like a film director: bird's eye, worm's eye, close-up, wide establishing shot. This creates dynamic visual storytelling.

Abstract Art for Wall Decor

29/48

✨ What it does

ChatGPT generates print-ready abstract art in a [geometric / organic / expressionist / minimalist / color field] style, using colors that fit the room you describe. You generate a few and pick the piece you would actually hang.

Generate an abstract art piece suitable for printing and framing as wall decor. Style: [geometric / organic / expressionist / minimalist / color field]. Dominant colors: [list 3-5 colors that would complement a room with (describe room colors)]. The piece should evoke [describe mood: calm, energetic, contemplative, luxurious]. Composition: [balanced / asymmetric / radial / flowing]. Include [texture: visible brushstrokes, palette knife marks, smooth gradients, splatter]. Dimensions: [landscape / portrait / square] orientation. The result should look like a piece you would find in a [contemporary gallery / boutique hotel / Scandinavian home / corporate lobby].

💡

Pro tip: Pull your color palette from the room where you will hang the art. A piece that shares 2-3 colors with its environment looks curated, not random.

Holiday & Event Photos

7 prompts

Christmas Card Photo

30/48

✨ What it does

ChatGPT generates a Christmas card photo of [describe people: family of four, couple, single person with pet] in a holiday setting, posed the way you ask. You generate a few and pick the one you would print.

Generate a Christmas card photo featuring [describe people: family of four, couple, single person with pet]. Physical descriptions: [detail each person]. Setting: [in front of a Christmas tree, on a snowy porch, by a fireplace with stockings, in matching holiday pajamas]. Everyone is [posing description]. Include Christmas elements: [twinkling lights, ornaments, wreaths, hot cocoa, presents]. The mood should be [warm and cozy / playful and fun / elegant and classic]. Lighting: [warm indoor / soft golden / sparkly]. Leave space at [top/bottom] for text that will read "[Family name] Christmas [year]". Aspect ratio: 5x7 vertical for printing.

💡

Pro tip: Generate the image in early November to have time for printing and mailing. Rush jobs in December mean fewer card options.

Birthday Party Invitation Visual

31/48

✨ What it does

ChatGPT generates a birthday invitation image for an [age]-year-old, in a [describe theme: tropical, princess, superhero, elegant adult dinner, minimalist modern] look, with empty space for name, date, and time. You generate a few and pick one, then add the details.

Generate a birthday party invitation image for a [age]-year-old's birthday. Theme: [describe theme: tropical, princess, superhero, elegant adult dinner, minimalist modern]. The image should include: [visual elements for the theme]. Color palette: [list colors]. Leave clear space for these text elements: Name, date, time, location, RSVP info. The layout should be [portrait 5x7 / landscape / square for digital]. Style: [playful illustrated / photo-realistic / graphic design / hand-lettered]. Include [balloons, confetti, streamers, flowers, themed decorations] as appropriate for the theme.

💡

Pro tip: For kids' parties, make the theme elements BIG and obvious. For adult parties, subtlety and color palette carry the theme.

Wedding or Engagement Announcement

32/48

✨ What it does

ChatGPT generates a [wedding invitation / engagement announcement / save the date] image of the couple in a [classic and elegant / modern minimalist / rustic bohemian / tropical destination / romantic garden] setting. You generate a few and pick the one that matches the wedding.

Generate an image for a [wedding invitation / engagement announcement / save the date]. The couple: [describe both people]. Style: [classic and elegant / modern minimalist / rustic bohemian / tropical destination / romantic garden]. Setting: [describe background]. The couple is [posing description]. Color palette: [wedding colors]. Include [floral elements, arch, venue details, seasonal elements]. Leave space for text: names, date, venue, and [other details]. Aspect ratio: [5x7 for printed invitations / 1:1 for social media announcement / 16:9 for wedding website header]. The overall aesthetic should match a [price range / venue type] wedding.

💡

Pro tip: Match the image style to your actual wedding venue and decor plan. Guests form expectations from the invitation.

Halloween Costume Concept

33/48

✨ What it does

ChatGPT generates a Halloween costume photo of a [physical description] person dressed as [costume idea], with the clothes, makeup, and props you list. You generate a few and pick the look before you buy or sew anything.

Generate a Halloween costume concept photo showing a person: [physical description] dressed as [costume idea]. The costume includes: [describe each element in detail: clothing, makeup, props, accessories]. Setting: [describe background: dark atmospheric, trick-or-treat street, haunted house, costume party]. The photo style should be [spooky and atmospheric / fun and colorful / cinematic horror / cute and family-friendly]. Lighting: [dramatic with harsh shadows / orange jack-o-lantern glow / purple-blue moody / bright party lighting]. Include [jack-o-lanterns, cobwebs, fog, candy, bats] as set dressing.

💡

Pro tip: Generate the costume concept in September. This gives you time to source materials, test makeup looks, and make adjustments before October 31.

Graduation Photo

34/48

✨ What it does

ChatGPT generates a graduation photo of [person description] in [describe cap and gown: color, stole, honor cords], on campus or under an arch, in the pose you choose. You generate a few and pick one to share or frame.

Generate a graduation photo of [person description] wearing [describe cap and gown: color, stole, honor cords]. Setting: [university campus steps, library, under a brick archway, with flowering trees]. The graduate is [posing: tossing cap, holding diploma, standing proudly, laughing with friends]. Include [university-specific details if any]. The mood is [celebratory / reflective / proud / joyful]. Lighting: [golden hour outdoor / bright spring day / formal indoor]. Aspect ratio: [portrait for framing / square for social media]. Style: editorial graduation photography.

💡

Pro tip: Include the specific gown and stole colors for your university. Generic black gowns feel stock-photo-like, while specific details make it personal.

Event Promotional Graphic

35/48

✨ What it does

ChatGPT generates a promo graphic for [event name], a [type of event: conference, workshop, concert, charity gala, product launch, webinar] on [date], with space for the key details. You generate a few and pick the one that matches the mood.

Generate a promotional graphic for [event name], a [type of event: conference, workshop, concert, charity gala, product launch, webinar]. Date: [date]. The visual should communicate: [the theme or mood of the event]. Include: [relevant visual elements - stage, speakers, products, venue]. Color palette: [event brand colors]. Text areas needed for: event name (large), date and time, location, tagline or brief description, and a CTA. Aspect ratio: [16:9 for event page banner / 1:1 for social media / 9:16 for Story promotion]. Style: [corporate professional / creative and bold / elegant and upscale / casual and inviting].

💡

Pro tip: Create the graphic in multiple aspect ratios from the same visual concept. You will need different versions for different channels.

Seasonal Business Promotion Visual

36/48

✨ What it does

ChatGPT generates a seasonal promo for [business name], a [business type], for [season/holiday: Valentine's Day, Summer Sale, Back to School, Black Friday, New Year], mixing brand colors with holiday cues. You generate a few and pick the one you would run.

Generate a seasonal promotional image for [business name], a [business type], for [season/holiday: Valentine's Day, Summer Sale, Back to School, Black Friday, New Year]. The image should feature [products or service visual]. Seasonal elements: [holiday-specific decorations, colors, symbols]. Brand colors: [list] should be prominently featured alongside seasonal colors. Include clear space for a promotional message like "[discount/offer]". The image should feel [festive but professional / fun and eye-catching / elegant and seasonal / urgent and promotional]. Aspect ratio: [specify for intended use]. Target audience: [describe].

💡

Pro tip: Your brand colors should still dominate even in seasonal promotions. A sale that does not look like your brand confuses returning customers.

Free tool

Image Prompt Builder

Build a rich copy-paste prompt for Midjourney, DALL-E, or Stable Diffusion.

Try it free →

Frequently Asked Questions

ChatGPT can generate images based on your physical description, but they will not be exact replicas of your face. For the most accurate results, provide extremely detailed descriptions including face shape, specific features, and distinguishing characteristics. The output works well for conceptual visuals, social media content, and creative projects, but should not be used as a substitute for actual photos in contexts where authenticity matters, like professional ID photos or legal documents.
Images generated through ChatGPT (DALL-E) are generally yours to use commercially per OpenAI's terms of service as of 2026. However, you should avoid generating images that closely resemble real, identifiable people or copyrighted characters. Always check the current terms of service for the specific AI tool you use, as policies can change. For any use in advertising or products, it is good practice to disclose that images are AI-generated where required by applicable regulations.
Save your detailed character or product description as a reusable block that you paste into every prompt. Include specific details: exact hair color and style, clothing, body proportions, and distinguishing features. Reference previous outputs by describing them ("maintain the same face shape and expression from the previous image"). Consistency improves significantly when your descriptions are specific enough that two different AI sessions would produce similar results from the same prompt.
The best prompts include five elements: subject (who or what), setting (where), style (how it looks), lighting (the mood driver), and composition (camera angle and framing). Vague prompts like "a photo of a dog" produce generic results. Specific prompts like "a golden retriever puppy sitting on a worn leather armchair, warm afternoon light through a bay window, shot with a 50mm lens at f/1.8, shallow depth of field" produce images you can actually use.
Yes, these prompts work with any text-to-image AI including Midjourney, DALL-E, Stable Diffusion, and Adobe Firefly. Each tool has different strengths: Midjourney excels at artistic and stylized images, DALL-E is strong at following complex instructions, and Stable Diffusion offers the most control over technical parameters. You may need to adjust formatting slightly (Midjourney uses different syntax), but the descriptive content of each prompt is universally applicable.