AI image prompts for Midjourney, ChatGPT, Gemini and Flux
These free AI image prompts are fill-in templates for text-to-image tools such as Midjourney, ChatGPT (DALL·E), Google Gemini, Flux and Stable Diffusion. Each one names the subject, setting, style, lighting, composition and aspect ratio, and lists what to avoid, which is what image models need to get close on the first try. Copy a prompt, fill in the [BLANKS] and paste it into your tool.
By the ThinkWithPrompt editorial team at Synottic · 26 prompts · Updated October 2026
What makes a good AI image prompt?
- Subject
- What is in the picture, with specific details: “a ceramic mug with a speckled glaze”, not “a mug”.
- Setting
- Where it is and what is behind it.
- Style or medium
- Photograph, flat vector, watercolour, 3D render, gouache, line art.
- Lighting and colour
- Soft window light, golden hour, studio light; a palette of 3–5 colours.
- Composition
- Camera angle, framing and where to leave empty space for text.
- Aspect ratio
- 1:1 for marketplaces and feeds, 4:5 for Instagram, 16:9 for headers and slides, 9:16 for stories.
- What to avoid
- Text, logos, extra fingers, clutter. Most tools follow a short “avoid” line.
How do you use these prompts in each tool?
- Midjourney
- Paste the prompt, then set the aspect ratio with its own parameter (for example --ar 16:9). Shorter prompts often work better; use the short version the first prompt below generates.
- ChatGPT (DALL·E / GPT image)
- Paste the full prompt in a chat. You can ask for changes in plain language afterwards, such as “same image, warmer light”.
- Google Gemini
- Paste the prompt in the Gemini app and ask it to create an image. To edit one of your own photos instead, see the photo prompts.
- Flux and Stable Diffusion
- Put the most important details first and move the “avoid” line into the negative prompt field if your tool has one.
12 AI image prompts to copy and paste
Highlighted [brackets] are blanks: replace them with your own details before sending.
Turns a one-line picture idea into a detailed image prompt with subject, style, lighting and composition, plus versions for different tools.
Act as an expert prompt writer for AI image generators.
My idea: [ONE OR TWO SENTENCES]
Tool I'll use: [Midjourney / ChatGPT (DALL·E) / Gemini / Flux / Stable Diffusion / other]
Where the image will be used: [website hero / social post / slide / print / just for fun]
Aspect ratio: [16:9 / 1:1 / 4:5 / 9:16]
Task: Write one detailed image prompt that describes, in this order:
1. Subject and action, with specific visual details.
2. Setting and background.
3. Style or medium (photo, illustration, 3D render, watercolour…).
4. Lighting and colour palette.
5. Composition and camera angle.
6. Mood.
Then give: a short version for tools that prefer brief prompts, two variations with a different style, and a list of things to avoid (for example extra fingers, text, clutter).
Rules: no real people, named living artists, brands or copyrighted characters.
A text-to-image template for a clean e-commerce product shot on a plain background with soft studio light.
Create a professional studio product photograph.
Product: [describe the product: shape, colour, material, size]
Background: [pure white / soft grey gradient / pastel colour: COLOUR]
Angle: [front / three-quarter view / top-down flat lay]
Props (optional): [none / a few simple props that suit the product]
Aspect ratio: [1:1 for marketplaces / 4:5 / 16:9]
Style: high-end commercial product photography, soft diffused studio lighting, gentle realistic shadow under the product, sharp focus across the whole product, true-to-life colours, crisp edges, lots of clean space around the product.
Avoid: text, logos, watermarks, reflections that hide the product, distorted shapes, extra objects.
Note: an AI image is a concept or mock-up; use real photos of your actual product for listings so buyers see exactly what they get.
A prompt for an AI lifestyle image that shows a product being used in a realistic setting, for ads and social posts.
Create a lifestyle photograph showing a product in use.
Product: [describe it]
Setting: [cosy kitchen in morning light / modern office desk / outdoor café / bathroom shelf / hiking trail]
Who is using it (optional): [no people / hands only / a person described generally, e.g. a woman in her 30s, seen from behind]
Mood: [calm / energetic / premium / homely]
Colour palette: [warm neutrals / bright and fresh / brand colours: COLOURS]
Aspect ratio: [4:5 / 1:1 / 9:16 / 16:9]
Style: natural editorial photography, soft natural light, shallow depth of field with the product in sharp focus, realistic textures, uncluttered composition with space for text on one side.
Avoid: text, logos, distorted hands, extra fingers, identifiable real people.
A prompt for exploring simple logo concepts with an AI image tool, as inspiration for a designer to refine into a real logo.
Create a sheet of 6 simple logo concepts.
Brand name: [NAME]
What the business does: [ONE LINE]
Personality: [e.g. trustworthy, playful, premium, eco-friendly, technical]
Symbol ideas (optional): [e.g. a leaf, a bridge, an abstract mark]
Colours: [up to 2–3 colours, or black only]
Style: flat vector logo marks, simple geometric shapes, strong silhouette that works small and in one colour, centred on a plain white background, each concept in its own square, evenly spaced grid.
Avoid: gradients, photorealism, tiny details, mock-ups, resemblance to existing well-known logos.
Note: AI image tools often misspell text, so ask for symbols only and add the brand name in design software. Use these as concepts for a designer, and check trademarks before using any logo.
A prompt for clean flat vector illustrations for websites, blog headers and presentations, in a consistent style.
Create a flat vector illustration.
Scene or concept: [e.g. a small team planning around a whiteboard / a person working remotely from a balcony / data flowing between devices]
Use: [website hero / blog header / slide / empty-state graphic]
Colour palette: [3–5 colours: COLOURS, or a calm blue and coral palette]
Background: [plain / soft shapes / white]
Aspect ratio: [16:9 / 4:3 / 1:1]
Style: modern flat vector illustration, simple geometric shapes, consistent line weight, limited colour palette, friendly people with simplified faces, generous negative space, clean edges suitable for the web.
Avoid: text, logos, photorealism, clutter, gradients unless I ask for them.
Tip: to keep a set consistent, reuse the same style sentence and palette for every illustration.
A prompt for soft watercolour-style illustrations for cards, prints, book covers and social posts.
Create a watercolour illustration.
Subject: [e.g. a quiet harbour at dawn / a bouquet of wildflowers in a jar / a fox in a snowy forest]
Colour palette: [soft pastels / muted earth tones / vibrant: COLOURS]
Composition: [centred subject with white space / full scene / vignette that fades to white edges]
Use: [greeting card / wall print / book cover / social post]
Aspect ratio: [5:7 / 1:1 / 4:5]
Style: traditional watercolour on textured cold-press paper, soft wet-on-wet blends, visible pigment granulation, delicate pencil underdrawing, gentle light, hand-painted feel.
Avoid: text, digital-looking hard edges, oversaturation, imitation of any named living artist.
A prompt for clean 3D isometric icons and mini-scenes for apps, websites and presentations.
Create a 3D isometric illustration.
Object or scene: [e.g. a tiny office with a desk and plants / a shield with a lock for security / a shopping cart with parcels]
Use: [app icon / feature icon set / landing page / slide]
Colour palette: [soft pastel / brand colours: COLOURS]
Background: [plain white / soft gradient / transparent-looking light grey]
Aspect ratio: [1:1 / 4:3]
Style: clean 3D isometric render, soft rounded shapes, matte clay-like materials, soft studio lighting with gentle shadows, minimal detail, centred composition with space around it.
Avoid: text, logos, clutter, photorealistic textures.
Tip: for an icon set, keep the same lighting, angle and palette sentence for every icon and change only the object.
A prompt for dramatic, cinematic landscape and scene images for wallpapers, headers and storytelling.
Create a cinematic wide-angle image.
Scene: [e.g. a lone lighthouse on cliffs in a storm / a neon-lit city street after rain / misty mountains at sunrise / a desert road at dusk]
Time and weather: [golden hour / blue hour / fog / rain / clear night with stars]
Mood: [epic / peaceful / mysterious / hopeful]
Aspect ratio: [21:9 / 16:9 / 9:16 for phone wallpaper]
Style: cinematic photography, wide-angle lens, strong sense of depth with foreground, middle ground and background, dramatic natural lighting, rich but realistic colour grading, atmospheric haze, high detail.
Avoid: text, watermarks, people close to the camera, oversaturated colours.
Variation: add "shot on 35mm film, subtle grain" for a more analogue look.
A prompt for appetising food images for menus, recipe blogs and social posts, with styling and lighting control.
Create a food photograph.
Dish: [describe the dish, ingredients and garnish]
Plate and surface: [rustic wooden board / white ceramic plate on linen / dark slate]
Angle: [overhead flat lay / 45-degree / close-up at table height]
Props: [cutlery, napkin, a few raw ingredients, a drink]
Lighting: [soft window light from the side / moody dark background with one light]
Aspect ratio: [4:5 / 1:1 / 16:9]
Style: professional editorial food photography, appetising and fresh, realistic textures, gentle steam if hot, shallow depth of field, natural colours.
Avoid: text, logos, plastic-looking food, extra cutlery or hands in odd positions.
Note: if you sell this dish, use a real photo in menus and listings so customers see what they'll get.
A prompt for visualising a room design idea, style and colour scheme before you renovate or decorate.
Create an interior design visualisation.
Room: [living room / bedroom / kitchen / home office / café interior]
Size and shape: [e.g. small square room with one large window]
Style: [Scandinavian / Japandi / industrial / mid-century modern / traditional Indian / coastal]
Colour palette and materials: [e.g. warm white walls, oak wood, linen, terracotta accents]
Key furniture and features: [LIST]
Aspect ratio: [16:9 / 4:3]
Style: photorealistic architectural interior photography, natural daylight from the window, wide-angle lens at eye level, straight vertical lines, realistic materials and proportions, styled but lived-in.
Avoid: text, distorted furniture, impossible structures, people.
Note: this is a mood visual; check real dimensions, materials and safety with a professional before building.
A prompt for warm, consistent children's picture-book illustrations with a recurring character you describe.
Create a children's picture-book illustration.
Main character (keep identical in every image): [e.g. a small round hedgehog with a red scarf and big curious eyes]
Scene: [what is happening on this page]
Setting: [e.g. a cosy village bakery in winter]
Mood: [cosy / adventurous / funny / calm bedtime]
Aspect ratio: [4:3 / 1:1 / two-page spread 2:1]
Style: warm hand-painted gouache look, soft rounded shapes, gentle textures, friendly expressions, rich but soft colours, simple background that leaves space for text at the [top / bottom].
Avoid: text in the image, scary or sharp features, copyrighted characters, imitation of any named living illustrator.
Tip: paste the same character description into every page's prompt, and upload a previous image as a reference if your tool allows it.
A prompt for seamless, repeating patterns for fabric, wrapping paper, wallpapers and website backgrounds.
Create a seamless repeating pattern.
Motifs: [e.g. lemons and leaves / geometric tiles / tiny stars and moons / paisley]
Style: [flat vector / watercolour / block print / line art]
Colours: [background COLOUR plus 2–4 motif colours]
Density: [sparse / medium / dense]
Use: [fabric / gift wrap / phone wallpaper / website background]
Style: seamless tileable pattern, edges that continue perfectly when repeated, evenly balanced motifs with no obvious grid, flat even lighting, high resolution, square format 1:1.
Avoid: text, logos, shadows or perspective, a single large central object.
Tip: test the tile by placing four copies side by side; if a seam shows, ask for the edges to be fixed.
More image and design prompts
- Write a detailed prompt for an AI image generatorTurns a rough idea into a precise prompt for Midjourney, DALL·E, Imagen, Firefly, or Stable Diffusion.
- Create an image in a signature styleReady-to-fill templates for popular looks: double exposure, film photography, cinematic, collage, sketch, vintage illustration, and more.
- Create a 3D isometric miniature of a city or landmarkThe popular tiny-world diorama look for any city, landmark, or scene, with weather and lighting.
- Design stickers, T-shirts, and merch graphicsPrint-ready design concepts for stickers, apparel, and merchandise.
- Create coloring book pagesThemed coloring-page concepts and line-art prompts for kids or adults.
- Create a consistent character reference sheetFront, side, and back views, expressions, and details to keep a character consistent across images.
- Create a scientific or academic figurePublication-style diagrams and figures for papers, posters, and teaching.
- Design a costume or uniform for a film, game, or storyA costume design brief with period, function, materials, color, and insignia logic, plus prompts for concept art.
- Write AI prompts for virtually staging a real estate listingRoom-by-room virtual staging prompts that keep the real room unchanged, match the buyer profile, and meet disclosure rules for listings.
- Design a logo conceptLogo directions with rationale, plus prompts or briefs to produce them.
- Design a book coverCover concepts that signal genre, with title treatment and an image prompt.
- Design a poster, flyer, or event graphicLayout, typography, imagery, and copy for posters, flyers, and event graphics — or an AI image prompt to create one.
- Create an icon or illustration as SVG codeClean, scalable SVG graphics written as code.
- Design an app icon or icon setApp icons and consistent icon sets with style rules and prompts.
Frequently asked questions
What is a good AI image prompt?
A good image prompt is specific about the subject, setting, style, lighting, composition and aspect ratio, and says what to avoid. “A product photo of a white ceramic mug on a pale grey background, soft studio light, three-quarter view, 1:1, no text” works far better than “a nice mug photo”.
Do these prompts work in Midjourney?
Yes. Paste the prompt and set the aspect ratio with Midjourney’s own parameter. If the result drifts, shorten the prompt to the subject, style and lighting.
Can AI images include text?
Newer models handle short text better, but spelling still goes wrong often. Keep text out of the image and add it in a design tool, or ask for one short word and check it carefully.
Can I use AI images commercially?
It depends on the tool’s terms and your plan. Avoid real people, brands, logos and copyrighted characters, don’t imitate named living artists, and check the tool’s licence before using images in products or ads.
How do I edit my own photo with AI instead?
Use the AI photo editing prompts: upload your photo to Gemini or ChatGPT and the prompt keeps faces exactly as they are.