Gemini prompts for image generation

A prompt for Gemini image generation should describe the scene in full sentences: subject, setting, light, camera or art style, composition and aspect ratio.

Steaming stone-grey ceramic mug on a worn oak table in warm morning light
Small red cabin with one lit window beside a still fjord at blue hour
Isometric cutaway of a small corner bakery with oven, counter and shelves of bread

Image pack25 image prompts

25 Gemini image generation prompts to copy

Show every prompt for

Style

Light

Ratio

Steaming stone-grey ceramic mug on a worn oak table in warm morning light

Product shot of a ceramic mug

Negative space and a named light make this reusable as an ad or a blog hero.

Any model

A matte stone-grey ceramic mug with a thick handle, steam curling from fresh coffee. Setting: a worn oak table beside a linen napkin and a small sprig of dried lavender, a sunny kitchen softly blurred behind. Light: low morning sun from the left with long soft shadows and a gentle fill on the right. Camera: macro lens, moderate depth of field, crisp glaze texture, shallow depth of field. Composition: eye-level, the mug on the lower right third, negative space to the upper left for text. Mood: calm and inviting. Format: 4:5. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A matte stone-grey ceramic mug with a thick handle, steam curling from fresh coffee. Setting: worn oak table beside a linen napkin and a small sprig of dried lavender, a sunny kitchen softly blurred behind. Lighting: low morning sun from the left with long soft shadows and a gentle fill on the right. Camera: macro lens, moderate depth of field, crisp glaze texture, shallow depth of field. Composition: eye-level, the mug on the lower right third, negative space to the upper left for text. Keep the mood calm and inviting. Use a tall portrait frame; if it can't match 4:5 exactly, compose so I can crop to 4:5. No text, logos or watermarks.

Gemini

Create a photorealistic image. A matte stone-grey ceramic mug with a thick handle, steam curling from fresh coffee. Setting: worn oak table beside a linen napkin and a small sprig of dried lavender, a sunny kitchen softly blurred behind. Lighting: low morning sun from the left with long soft shadows and a gentle fill on the right. Camera: macro lens, moderate depth of field, crisp glaze texture, shallow depth of field. Composition: eye-level, the mug on the lower right third, negative space to the upper left for text. Keep the mood calm and inviting. Make it a 4:5 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A matte stone-grey ceramic mug with a thick handle, steam curling from fresh coffee. Setting: worn oak table beside a linen napkin and a small sprig of dried lavender, a sunny kitchen softly blurred behind. Lighting: low morning sun from the left with long soft shadows and a gentle fill on the right. Camera: macro lens, moderate depth of field, crisp glaze texture, shallow depth of field. Composition: eye-level, the mug on the lower right third, negative space to the upper left for text. Keep the mood calm and inviting. Aspect ratio 4:5. No text, logos or watermarks.

Midjourney

A matte stone-grey ceramic mug with a thick handle, steam curling from fresh coffee, a worn oak table beside a linen napkin and a small sprig of dried lavender, a sunny kitchen softly blurred behind, low morning sun from the left with long soft shadows and a gentle fill on the right, macro lens, moderate depth of field, crisp glaze texture, shallow depth of field, eye-level, the mug on the lower right third, negative space to the upper left for text, calm and inviting --ar 4:5 --no text, logo, watermark --raw

How to use
  1. Copy it with the button above.
  2. Fill the blanks in {{double braces}} with your own details.
  3. Paste it into your image AI and generate.
  4. Then ask for one change at a time, not a new prompt.

Before you run “Product shot of a ceramic mug” for the first time, the lesson on how to use AI image prompts walks through copying, filling and sending it.

Overhead ramen bowl flat lay

A diagonal prop layout stops an overhead shot looking static.

Any model

A deep black bowl of tonkotsu ramen with a jammy egg, chashu slices, nori and spring onion, steam rising. Setting: a dark wooden counter with chopsticks on a rest, a small dish of chilli oil and a folded napkin. Light: soft window light from above-left, a subtle glossy highlight on the broth. Camera: overhead at standard lens, moderate depth of field, sharp across the whole bowl. Composition: the bowl slightly off centre, props in a neat diagonal. Mood: warm and comforting. Format: 1:1. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A deep black bowl of tonkotsu ramen with a jammy egg, chashu slices, nori and spring onion, steam rising. Setting: dark wooden counter with chopsticks on a rest, a small dish of chilli oil and a folded napkin. Lighting: soft window light from above-left, a subtle glossy highlight on the broth. Camera: overhead at standard lens, moderate depth of field, sharp across the whole bowl. Composition: the bowl slightly off centre, props in a neat diagonal. Keep the mood warm and comforting. Use a square frame; if it can't match 1:1 exactly, compose so I can crop to 1:1. No text, logos or watermarks.

Gemini

Create a photorealistic image. A deep black bowl of tonkotsu ramen with a jammy egg, chashu slices, nori and spring onion, steam rising. Setting: dark wooden counter with chopsticks on a rest, a small dish of chilli oil and a folded napkin. Lighting: soft window light from above-left, a subtle glossy highlight on the broth. Camera: overhead at standard lens, moderate depth of field, sharp across the whole bowl. Composition: the bowl slightly off centre, props in a neat diagonal. Keep the mood warm and comforting. Make it a 1:1 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A deep black bowl of tonkotsu ramen with a jammy egg, chashu slices, nori and spring onion, steam rising. Setting: dark wooden counter with chopsticks on a rest, a small dish of chilli oil and a folded napkin. Lighting: soft window light from above-left, a subtle glossy highlight on the broth. Camera: overhead at standard lens, moderate depth of field, sharp across the whole bowl. Composition: the bowl slightly off centre, props in a neat diagonal. Keep the mood warm and comforting. Aspect ratio 1:1. No text, logos or watermarks.

Midjourney

A deep black bowl of tonkotsu ramen with a jammy egg, chashu slices, nori and spring onion, steam rising, a dark wooden counter with chopsticks on a rest, a small dish of chilli oil and a folded napkin, soft window light from above-left, a subtle glossy highlight on the broth, overhead at standard lens, moderate depth of field, sharp across the whole bowl, the bowl slightly off centre, props in a neat diagonal, warm and comforting --ar 1:1 --no text, logo, watermark --raw

Small red cabin with one lit window beside a still fjord at blue hour

Fjord at blue hour

One warm window in a cold scene creates a focal point without a person.

Any model

A narrow fjord between steep mountains at blue hour, a single small red cabin on the shoreline with one lit window. Setting: still water mirroring the peaks, low mist on the slopes, a few snow patches. Light: deep blue twilight with a thin pink glow on the highest ridge. Camera: ultra-wide lens, 20 second exposure look, silky water, tripod-sharp detail. Composition: the cabin on the lower left third, a diagonal shoreline leading to the mountains. Mood: quiet and remote. Format: 16:9. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A narrow fjord between steep mountains at blue hour, a single small red cabin on the shoreline with one lit window. Setting: still water mirroring the peaks, low mist on the slopes, a few snow patches. Lighting: deep blue twilight with a thin pink glow on the highest ridge. Camera: ultra-wide lens, 20 second exposure look, silky water, tripod-sharp detail. Composition: the cabin on the lower left third, a diagonal shoreline leading to the mountains. Keep the mood quiet and remote. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. No text, logos or watermarks.

Gemini

Create a photorealistic image. A narrow fjord between steep mountains at blue hour, a single small red cabin on the shoreline with one lit window. Setting: still water mirroring the peaks, low mist on the slopes, a few snow patches. Lighting: deep blue twilight with a thin pink glow on the highest ridge. Camera: ultra-wide lens, 20 second exposure look, silky water, tripod-sharp detail. Composition: the cabin on the lower left third, a diagonal shoreline leading to the mountains. Keep the mood quiet and remote. Make it a 16:9 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A narrow fjord between steep mountains at blue hour, a single small red cabin on the shoreline with one lit window. Setting: still water mirroring the peaks, low mist on the slopes, a few snow patches. Lighting: deep blue twilight with a thin pink glow on the highest ridge. Camera: ultra-wide lens, 20 second exposure look, silky water, tripod-sharp detail. Composition: the cabin on the lower left third, a diagonal shoreline leading to the mountains. Keep the mood quiet and remote. Aspect ratio 16:9. No text, logos or watermarks.

Midjourney

A narrow fjord between steep mountains at blue hour, a single small red cabin on the shoreline with one lit window, still water mirroring the peaks, low mist on the slopes, a few snow patches, deep blue twilight with a thin pink glow on the highest ridge, ultra-wide lens, 20 second exposure look, silky water, tripod-sharp detail, the cabin on the lower left third, a diagonal shoreline leading to the mountains, quiet and remote --ar 16:9 --no text, logo, watermark --raw

Scandinavian living room

Asking for straight verticals prevents the stretched look of wide interior shots.

Any model

A bright Scandinavian living room with a boucle sofa, a pale oak coffee table and a large leafy plant. Setting: white walls, a tall window with sheer curtains, a woven rug and a stack of art books. Light: diffused north daylight, soft shadows, no harsh highlights. Camera: ultra-wide interior lens, vertical lines straight, deep focus. Composition: a corner view with the sofa on the left and the window light falling across the rug. Mood: airy and calm. Format: 3:2. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A bright Scandinavian living room with a boucle sofa, a pale oak coffee table and a large leafy plant. Setting: white walls, a tall window with sheer curtains, a woven rug and a stack of art books. Lighting: diffused north daylight, soft shadows, no harsh highlights. Camera: ultra-wide interior lens, vertical lines straight, deep focus. Composition: a corner view with the sofa on the left and the window light falling across the rug. Keep the mood airy and calm. Use a landscape frame; if it can't match 3:2 exactly, compose so I can crop to 3:2. No text, logos or watermarks.

Gemini

Create a photorealistic image. A bright Scandinavian living room with a boucle sofa, a pale oak coffee table and a large leafy plant. Setting: white walls, a tall window with sheer curtains, a woven rug and a stack of art books. Lighting: diffused north daylight, soft shadows, no harsh highlights. Camera: ultra-wide interior lens, vertical lines straight, deep focus. Composition: a corner view with the sofa on the left and the window light falling across the rug. Keep the mood airy and calm. Make it a 3:2 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A bright Scandinavian living room with a boucle sofa, a pale oak coffee table and a large leafy plant. Setting: white walls, a tall window with sheer curtains, a woven rug and a stack of art books. Lighting: diffused north daylight, soft shadows, no harsh highlights. Camera: ultra-wide interior lens, vertical lines straight, deep focus. Composition: a corner view with the sofa on the left and the window light falling across the rug. Keep the mood airy and calm. Aspect ratio 3:2. No text, logos or watermarks.

Midjourney

A bright Scandinavian living room with a boucle sofa, a pale oak coffee table and a large leafy plant, white walls, a tall window with sheer curtains, a woven rug and a stack of art books, diffused north daylight, soft shadows, no harsh highlights, ultra-wide interior lens, vertical lines straight, deep focus, a corner view with the sofa on the left and the window light falling across the rug, airy and calm --ar 3:2 --no text, logo, watermark --raw

Isometric cutaway of a small corner bakery with oven, counter and shelves of bread

Isometric cutaway of a tiny bakery

Naming a single floating room keeps the image tidy and prevents a messy city scene.

Any model

An isometric 3D cutaway of a small corner bakery with an oven, a display counter, shelves of bread and a tiny baker. Style: clean isometric 3D render, soft clay-like materials, rounded edges. Palette: warm peach, cream and terracotta with one teal accent. Light: soft global illumination with gentle shadows on a plain pale background. Composition: a single floating cube room viewed from the corner at 45 degrees. Format: 1:1. No text, logos or watermarks.

ChatGPT

Create an image. An isometric 3D cutaway of a small corner bakery with an oven, a display counter, shelves of bread and a tiny baker. Style: clean isometric 3D render, soft clay-like materials, rounded edges. Palette: warm peach, cream and terracotta with one teal accent. Lighting: soft global illumination with gentle shadows on a plain pale background. Composition: a single floating cube room viewed from the corner at 45 degrees. Use a square frame; if it can't match 1:1 exactly, compose so I can crop to 1:1. No text, logos or watermarks.

Gemini

Create an image. An isometric 3D cutaway of a small corner bakery with an oven, a display counter, shelves of bread and a tiny baker. Style: clean isometric 3D render, soft clay-like materials, rounded edges. Palette: warm peach, cream and terracotta with one teal accent. Lighting: soft global illumination with gentle shadows on a plain pale background. Composition: a single floating cube room viewed from the corner at 45 degrees. Make it a 1:1 frame. No text, logos or watermarks.

Nano Banana

Generate an image. An isometric 3D cutaway of a small corner bakery with an oven, a display counter, shelves of bread and a tiny baker. Style: clean isometric 3D render, soft clay-like materials, rounded edges. Palette: warm peach, cream and terracotta with one teal accent. Lighting: soft global illumination with gentle shadows on a plain pale background. Composition: a single floating cube room viewed from the corner at 45 degrees. Aspect ratio 1:1. No text, logos or watermarks.

Midjourney

An isometric 3D cutaway of a small corner bakery with an oven, a display counter, shelves of bread and a tiny baker, clean isometric 3D render, soft clay-like materials, rounded edges, warm peach, cream and terracotta with one teal accent, soft global illumination with gentle shadows on a plain pale background, a single floating cube room viewed from the corner at 45 degrees --ar 1:1 --no text, logo, watermark --raw

Watercolour of a red fox among ferns at the edge of a misty birch forest

Watercolour fox in a misty forest

Describing paper and pigment behaviour gives the AI a real medium to copy.

Any model

A red fox sitting among ferns at the edge of a misty birch forest, looking back over its shoulder. Style: loose watercolour on rough paper with visible pigment blooms and white paper left untouched. Palette: soft greens, russet and warm grey. Light: pale morning light diffusing through mist. Composition: the fox in the lower right, tree trunks fading into pale washes in the upper left. Format: 4:5. No text, logos or watermarks.

ChatGPT

Create an image. A red fox sitting among ferns at the edge of a misty birch forest, looking back over its shoulder. Style: loose watercolour on rough paper with visible pigment blooms and white paper left untouched. Palette: soft greens, russet and warm grey. Lighting: pale morning light diffusing through mist. Composition: the fox in the lower right, tree trunks fading into pale washes in the upper left. Use a tall portrait frame; if it can't match 4:5 exactly, compose so I can crop to 4:5. No text, logos or watermarks.

Gemini

Create an image. A red fox sitting among ferns at the edge of a misty birch forest, looking back over its shoulder. Style: loose watercolour on rough paper with visible pigment blooms and white paper left untouched. Palette: soft greens, russet and warm grey. Lighting: pale morning light diffusing through mist. Composition: the fox in the lower right, tree trunks fading into pale washes in the upper left. Make it a 4:5 frame. No text, logos or watermarks.

Nano Banana

Generate an image. A red fox sitting among ferns at the edge of a misty birch forest, looking back over its shoulder. Style: loose watercolour on rough paper with visible pigment blooms and white paper left untouched. Palette: soft greens, russet and warm grey. Lighting: pale morning light diffusing through mist. Composition: the fox in the lower right, tree trunks fading into pale washes in the upper left. Aspect ratio 4:5. No text, logos or watermarks.

Midjourney

A red fox sitting among ferns at the edge of a misty birch forest, looking back over its shoulder, loose watercolour on rough paper with visible pigment blooms and white paper left untouched, soft greens, russet and warm grey, pale morning light diffusing through mist, the fox in the lower right, tree trunks fading into pale washes in the upper left --ar 4:5 --no text, logo, watermark --raw

Rainy cyberpunk night market

Invented script prevents real brand names or readable signs.

Any model

A crowded street food market at night in a futuristic city, steaming noodle stalls under holographic signs. Setting: wet streets, narrow alleys, hanging cables, vertical neon signs in invented script. Light: magenta and cyan neon reflecting on puddles, warm stall bulbs, volumetric haze. Camera: wide-angle cinematic lens, shallow depth of field, slight chromatic aberration. Composition: a low angle down the alley with a lone figure under an umbrella near the centre. Mood: busy, moody and alive. Format: 16:9. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A crowded street food market at night in a futuristic city, steaming noodle stalls under holographic signs. Setting: wet streets, narrow alleys, hanging cables, vertical neon signs in invented script. Lighting: magenta and cyan neon reflecting on puddles, warm stall bulbs, volumetric haze. Camera: wide-angle cinematic lens, shallow depth of field, slight chromatic aberration. Composition: a low angle down the alley with a lone figure under an umbrella near the centre. Keep the mood busy, moody and alive. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. No text, logos or watermarks.

Gemini

Create a photorealistic image. A crowded street food market at night in a futuristic city, steaming noodle stalls under holographic signs. Setting: wet streets, narrow alleys, hanging cables, vertical neon signs in invented script. Lighting: magenta and cyan neon reflecting on puddles, warm stall bulbs, volumetric haze. Camera: wide-angle cinematic lens, shallow depth of field, slight chromatic aberration. Composition: a low angle down the alley with a lone figure under an umbrella near the centre. Keep the mood busy, moody and alive. Make it a 16:9 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A crowded street food market at night in a futuristic city, steaming noodle stalls under holographic signs. Setting: wet streets, narrow alleys, hanging cables, vertical neon signs in invented script. Lighting: magenta and cyan neon reflecting on puddles, warm stall bulbs, volumetric haze. Camera: wide-angle cinematic lens, shallow depth of field, slight chromatic aberration. Composition: a low angle down the alley with a lone figure under an umbrella near the centre. Keep the mood busy, moody and alive. Aspect ratio 16:9. No text, logos or watermarks.

Midjourney

A crowded street food market at night in a futuristic city, steaming noodle stalls under holographic signs, wet streets, narrow alleys, hanging cables, vertical neon signs in invented script, magenta and cyan neon reflecting on puddles, warm stall bulbs, volumetric haze, wide-angle cinematic lens, shallow depth of field, slight chromatic aberration, a low angle down the alley with a lone figure under an umbrella near the centre, busy, moody and alive --ar 16:9 --no text, logo, watermark --raw

Vintage travel poster with text

Giving the exact text in capitals and quotes is the most reliable way to get readable lettering.

Any model

A mid-century style travel poster for {{city}} with a bold skyline and a large headline. Text: the headline reads VISIT {{city}} in tall condensed capitals, with a small line below reading "The Sunny Side of Travel". Style: flat vector illustration with grainy print texture and limited colours. Palette: sunset orange, deep teal, cream and navy. Composition: the skyline filling the lower half, a large sun behind it, the text at the top with generous margins. Format: 2:3. Spell the text exactly as written and check every letter.

ChatGPT

Create an image. A mid-century style travel poster for {{city}} with a bold skyline and a large headline. Text: the headline reads VISIT {{city}} in tall condensed capitals, with a small line below reading "The Sunny Side of Travel". Style: flat vector illustration with grainy print texture and limited colours. Palette: sunset orange, deep teal, cream and navy. Composition: the skyline filling the lower half, a large sun behind it, the text at the top with generous margins. Use a tall portrait frame; if it can't match 2:3 exactly, compose so I can crop to 2:3. Spell the text exactly as written and check every letter.

Gemini

Create an image. A mid-century style travel poster for {{city}} with a bold skyline and a large headline. Text: the headline reads VISIT {{city}} in tall condensed capitals, with a small line below reading "The Sunny Side of Travel". Style: flat vector illustration with grainy print texture and limited colours. Palette: sunset orange, deep teal, cream and navy. Composition: the skyline filling the lower half, a large sun behind it, the text at the top with generous margins. Make it a 2:3 frame. Spell the text exactly as written and check every letter.

Nano Banana

Generate an image. A mid-century style travel poster for {{city}} with a bold skyline and a large headline. Text: the headline reads VISIT {{city}} in tall condensed capitals, with a small line below reading "The Sunny Side of Travel". Style: flat vector illustration with grainy print texture and limited colours. Palette: sunset orange, deep teal, cream and navy. Composition: the skyline filling the lower half, a large sun behind it, the text at the top with generous margins. Aspect ratio 2:3. Spell the text exactly as written and check every letter.

Midjourney

A mid-century style travel poster for {{city}} with a bold skyline and a large headline, the headline reads VISIT {{city}} in tall condensed capitals, with a small line below reading "The Sunny Side of Travel", flat vector illustration with grainy print texture and limited colours, sunset orange, deep teal, cream and navy, the skyline filling the lower half, a large sun behind it, the text at the top with generous margins --ar 2:3 --raw

Macro dewdrop on a spider web

Asking for reflections inside the drops adds the detail that makes a macro shot feel real.

Any model

A spider web strung between two grass stems, covered in dew, each drop holding a tiny upside-down reflection of the garden. Setting: a garden at dawn with soft green bokeh behind. Light: low sunrise backlight making the drops sparkle. Camera: macro lens, shallow depth of field, focus stacked on the nearest strand. Composition: the web crossing the frame diagonally, the sharpest drops on the right third. Mood: delicate and still. Format: 3:2. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A spider web strung between two grass stems, covered in dew, each drop holding a tiny upside-down reflection of the garden. Setting: garden at dawn with soft green bokeh behind. Lighting: low sunrise backlight making the drops sparkle. Camera: macro lens, shallow depth of field, focus stacked on the nearest strand. Composition: the web crossing the frame diagonally, the sharpest drops on the right third. Keep the mood delicate and still. Use a landscape frame; if it can't match 3:2 exactly, compose so I can crop to 3:2. No text, logos or watermarks.

Gemini

Create a photorealistic image. A spider web strung between two grass stems, covered in dew, each drop holding a tiny upside-down reflection of the garden. Setting: garden at dawn with soft green bokeh behind. Lighting: low sunrise backlight making the drops sparkle. Camera: macro lens, shallow depth of field, focus stacked on the nearest strand. Composition: the web crossing the frame diagonally, the sharpest drops on the right third. Keep the mood delicate and still. Make it a 3:2 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A spider web strung between two grass stems, covered in dew, each drop holding a tiny upside-down reflection of the garden. Setting: garden at dawn with soft green bokeh behind. Lighting: low sunrise backlight making the drops sparkle. Camera: macro lens, shallow depth of field, focus stacked on the nearest strand. Composition: the web crossing the frame diagonally, the sharpest drops on the right third. Keep the mood delicate and still. Aspect ratio 3:2. No text, logos or watermarks.

Midjourney

A spider web strung between two grass stems, covered in dew, each drop holding a tiny upside-down reflection of the garden, a garden at dawn with soft green bokeh behind, low sunrise backlight making the drops sparkle, macro lens, shallow depth of field, focus stacked on the nearest strand, the web crossing the frame diagonally, the sharpest drops on the right third, delicate and still --ar 3:2 --no text, logo, watermark --raw

Brutalist library in fog

Fog is a cheap way to control depth and remove busy backgrounds.

Any model

A large concrete brutalist library with deep window recesses and a wide staircase, a few tiny people on the steps. Setting: a foggy morning in a quiet university square with bare trees. Light: flat, soft diffused fog light, muted tones, no strong shadows. Camera: wide-angle architecture lens, a slightly low angle, vertical lines straight. Composition: symmetrical, the stairs leading to the central entrance, the building fading into fog at the top. Mood: austere and quiet. Format: 4:5. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A large concrete brutalist library with deep window recesses and a wide staircase, a few tiny people on the steps. Setting: foggy morning in a quiet university square with bare trees. Lighting: flat, soft diffused fog light, muted tones, no strong shadows. Camera: wide-angle architecture lens, a slightly low angle, vertical lines straight. Composition: symmetrical, the stairs leading to the central entrance, the building fading into fog at the top. Keep the mood austere and quiet. Use a tall portrait frame; if it can't match 4:5 exactly, compose so I can crop to 4:5. No text, logos or watermarks.

Gemini

Create a photorealistic image. A large concrete brutalist library with deep window recesses and a wide staircase, a few tiny people on the steps. Setting: foggy morning in a quiet university square with bare trees. Lighting: flat, soft diffused fog light, muted tones, no strong shadows. Camera: wide-angle architecture lens, a slightly low angle, vertical lines straight. Composition: symmetrical, the stairs leading to the central entrance, the building fading into fog at the top. Keep the mood austere and quiet. Make it a 4:5 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A large concrete brutalist library with deep window recesses and a wide staircase, a few tiny people on the steps. Setting: foggy morning in a quiet university square with bare trees. Lighting: flat, soft diffused fog light, muted tones, no strong shadows. Camera: wide-angle architecture lens, a slightly low angle, vertical lines straight. Composition: symmetrical, the stairs leading to the central entrance, the building fading into fog at the top. Keep the mood austere and quiet. Aspect ratio 4:5. No text, logos or watermarks.

Midjourney

A large concrete brutalist library with deep window recesses and a wide staircase, a few tiny people on the steps, a foggy morning in a quiet university square with bare trees, flat, soft diffused fog light, muted tones, no strong shadows, wide-angle architecture lens, a slightly low angle, vertical lines straight, symmetrical, the stairs leading to the central entrance, the building fading into fog at the top, austere and quiet --ar 4:5 --no text, logo, watermark --raw

Houseplant sticker sheet

Counting the items and fixing the grid makes a usable sheet instead of a messy collage.

Any model

A sheet of twelve cute houseplant stickers: monstera, snake plant, cactus, fern, pothos, aloe and more, each in a small terracotta pot. Style: flat vector sticker style with thick white die-cut outlines and a tiny drop shadow. Palette: fresh greens, terracotta and soft cream. Composition: a neat 4 by 3 grid on a plain light grey background, equal spacing, every sticker complete. Format: 1:1. No text, logos or watermarks.

ChatGPT

Create an image. A sheet of twelve cute houseplant stickers: monstera, snake plant, cactus, fern, pothos, aloe and more, each in a small terracotta pot. Style: flat vector sticker style with thick white die-cut outlines and a tiny drop shadow. Palette: fresh greens, terracotta and soft cream. Composition: a neat 4 by 3 grid on a plain light grey background, equal spacing, every sticker complete. Use a square frame; if it can't match 1:1 exactly, compose so I can crop to 1:1. No text, logos or watermarks.

Gemini

Create an image. A sheet of twelve cute houseplant stickers: monstera, snake plant, cactus, fern, pothos, aloe and more, each in a small terracotta pot. Style: flat vector sticker style with thick white die-cut outlines and a tiny drop shadow. Palette: fresh greens, terracotta and soft cream. Composition: a neat 4 by 3 grid on a plain light grey background, equal spacing, every sticker complete. Make it a 1:1 frame. No text, logos or watermarks.

Nano Banana

Generate an image. A sheet of twelve cute houseplant stickers: monstera, snake plant, cactus, fern, pothos, aloe and more, each in a small terracotta pot. Style: flat vector sticker style with thick white die-cut outlines and a tiny drop shadow. Palette: fresh greens, terracotta and soft cream. Composition: a neat 4 by 3 grid on a plain light grey background, equal spacing, every sticker complete. Aspect ratio 1:1. No text, logos or watermarks.

Midjourney

A sheet of twelve cute houseplant stickers: monstera, snake plant, cactus, fern, pothos, aloe and more, each in a small terracotta pot, flat vector sticker style with thick white die-cut outlines and a tiny drop shadow, fresh greens, terracotta and soft cream, a neat 4 by 3 grid on a plain light grey background, equal spacing, every sticker complete --ar 1:1 --no text, logo, watermark --raw

Claymation bear baker

Visible fingerprints are what make clay look like clay.

Any model

A claymation bear in a tiny apron kneading dough, flour on its nose, in a miniature kitchen. Style: stop-motion claymation, visible fingerprints on the clay, handmade set pieces. Light: warm practical lamps with a soft shallow depth of field like a tabletop set. Camera: macro lens at eye level with the bear. Composition: the bear centre-left, a tiny oven glowing in the back right. Format: 3:2. No text, logos or watermarks.

ChatGPT

Create an image. A claymation bear in a tiny apron kneading dough, flour on its nose, in a miniature kitchen. Style: stop-motion claymation, visible fingerprints on the clay, handmade set pieces. Lighting: warm practical lamps with a soft shallow depth of field like a tabletop set. Camera: macro lens at eye level with the bear. Composition: the bear centre-left, a tiny oven glowing in the back right. Use a landscape frame; if it can't match 3:2 exactly, compose so I can crop to 3:2. No text, logos or watermarks.

Gemini

Create an image. A claymation bear in a tiny apron kneading dough, flour on its nose, in a miniature kitchen. Style: stop-motion claymation, visible fingerprints on the clay, handmade set pieces. Lighting: warm practical lamps with a soft shallow depth of field like a tabletop set. Camera: macro lens at eye level with the bear. Composition: the bear centre-left, a tiny oven glowing in the back right. Make it a 3:2 frame. No text, logos or watermarks.

Nano Banana

Generate an image. A claymation bear in a tiny apron kneading dough, flour on its nose, in a miniature kitchen. Style: stop-motion claymation, visible fingerprints on the clay, handmade set pieces. Lighting: warm practical lamps with a soft shallow depth of field like a tabletop set. Camera: macro lens at eye level with the bear. Composition: the bear centre-left, a tiny oven glowing in the back right. Aspect ratio 3:2. No text, logos or watermarks.

Midjourney

A claymation bear in a tiny apron kneading dough, flour on its nose, in a miniature kitchen, stop-motion claymation, visible fingerprints on the clay, handmade set pieces, warm practical lamps with a soft shallow depth of field like a tabletop set, macro lens at eye level with the bear, the bear centre-left, a tiny oven glowing in the back right --ar 3:2 --no text, logo, watermark --raw

Snow leopard in a blizzard

A telephoto lens and a catchlight are two cues for believable wildlife.

Any model

A snow leopard crouching on a rocky ledge in a snowstorm, frost on its whiskers, eyes fixed on something below. Setting: a high mountain ledge with swirling snow and rocks fading into white. Light: flat overcast light, muted blue-grey tones, a tiny catchlight in the eye. Camera: telephoto telephoto, moderate depth of field, shallow focus, snowflakes streaking slightly. Composition: the cat on the lower left third, snow filling the right. Mood: tense and wild. Format: 3:2. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A snow leopard crouching on a rocky ledge in a snowstorm, frost on its whiskers, eyes fixed on something below. Setting: high mountain ledge with swirling snow and rocks fading into white. Lighting: flat overcast light, muted blue-grey tones, a tiny catchlight in the eye. Camera: telephoto telephoto, moderate depth of field, shallow focus, snowflakes streaking slightly. Composition: the cat on the lower left third, snow filling the right. Keep the mood tense and wild. Use a landscape frame; if it can't match 3:2 exactly, compose so I can crop to 3:2. No text, logos or watermarks.

Gemini

Create a photorealistic image. A snow leopard crouching on a rocky ledge in a snowstorm, frost on its whiskers, eyes fixed on something below. Setting: high mountain ledge with swirling snow and rocks fading into white. Lighting: flat overcast light, muted blue-grey tones, a tiny catchlight in the eye. Camera: telephoto telephoto, moderate depth of field, shallow focus, snowflakes streaking slightly. Composition: the cat on the lower left third, snow filling the right. Keep the mood tense and wild. Make it a 3:2 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A snow leopard crouching on a rocky ledge in a snowstorm, frost on its whiskers, eyes fixed on something below. Setting: high mountain ledge with swirling snow and rocks fading into white. Lighting: flat overcast light, muted blue-grey tones, a tiny catchlight in the eye. Camera: telephoto telephoto, moderate depth of field, shallow focus, snowflakes streaking slightly. Composition: the cat on the lower left third, snow filling the right. Keep the mood tense and wild. Aspect ratio 3:2. No text, logos or watermarks.

Midjourney

A snow leopard crouching on a rocky ledge in a snowstorm, frost on its whiskers, eyes fixed on something below, a high mountain ledge with swirling snow and rocks fading into white, flat overcast light, muted blue-grey tones, a tiny catchlight in the eye, telephoto telephoto, moderate depth of field, shallow focus, snowflakes streaking slightly, the cat on the lower left third, snow filling the right, tense and wild --ar 3:2 --no text, logo, watermark --raw

Astronaut tending a plant on Mars

The suit wear and a small domestic action keep sci-fi grounded.

Any model

An astronaut in a worn white suit gently watering a tomato plant inside a glass greenhouse dome on Mars. Setting: a dusty red Martian landscape outside the dome, tools and trays inside. Light: warm dusk light through the dome glass and cool lamp light on the plants. Style: realistic concept art with fine suit wear and scratches. Composition: a medium shot from outside the dome, the astronaut framed by curved glass. Mood: hopeful and quiet. Format: 16:9. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. An astronaut in a worn white suit gently watering a tomato plant inside a glass greenhouse dome on Mars. Setting: dusty red Martian landscape outside the dome, tools and trays inside. Lighting: warm dusk light through the dome glass and cool lamp light on the plants. Style: realistic concept art with fine suit wear and scratches. Composition: a medium shot from outside the dome, the astronaut framed by curved glass. Keep the mood hopeful and quiet. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. No text, logos or watermarks.

Gemini

Create a photorealistic image. An astronaut in a worn white suit gently watering a tomato plant inside a glass greenhouse dome on Mars. Setting: dusty red Martian landscape outside the dome, tools and trays inside. Lighting: warm dusk light through the dome glass and cool lamp light on the plants. Style: realistic concept art with fine suit wear and scratches. Composition: a medium shot from outside the dome, the astronaut framed by curved glass. Keep the mood hopeful and quiet. Make it a 16:9 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. An astronaut in a worn white suit gently watering a tomato plant inside a glass greenhouse dome on Mars. Setting: dusty red Martian landscape outside the dome, tools and trays inside. Lighting: warm dusk light through the dome glass and cool lamp light on the plants. Style: realistic concept art with fine suit wear and scratches. Composition: a medium shot from outside the dome, the astronaut framed by curved glass. Keep the mood hopeful and quiet. Aspect ratio 16:9. No text, logos or watermarks.

Midjourney

An astronaut in a worn white suit gently watering a tomato plant inside a glass greenhouse dome on Mars, a dusty red Martian landscape outside the dome, tools and trays inside, warm dusk light through the dome glass and cool lamp light on the plants, realistic concept art with fine suit wear and scratches, a medium shot from outside the dome, the astronaut framed by curved glass, hopeful and quiet --ar 16:9 --no text, logo, watermark --raw

Lighthouse cross-section illustration

Short labels are likely to render correctly where long paragraphs fail.

Any model

A technical cutaway illustration of a lighthouse showing its spiral stairs, lamp room, keeper bedroom and engine room. Style: detailed vintage encyclopaedia illustration, fine ink lines and flat watercolour fills. Palette: navy, cream, rust red and soft yellow. Composition: the tower centred on a cream background, short labels with thin leader lines on both sides. Text: the labels read "Lamp room", "Spiral stairs", "Keeper bedroom" and "Engine room". Format: 2:3. Spell the labels exactly as written.

ChatGPT

Create an image. A technical cutaway illustration of a lighthouse showing its spiral stairs, lamp room, keeper bedroom and engine room. Style: detailed vintage encyclopaedia illustration, fine ink lines and flat watercolour fills. Palette: navy, cream, rust red and soft yellow. Composition: the tower centred on a cream background, short labels with thin leader lines on both sides. Text: the labels read "Lamp room", "Spiral stairs", "Keeper bedroom" and "Engine room". Use a tall portrait frame; if it can't match 2:3 exactly, compose so I can crop to 2:3. Spell the labels exactly as written.

Gemini

Create an image. A technical cutaway illustration of a lighthouse showing its spiral stairs, lamp room, keeper bedroom and engine room. Style: detailed vintage encyclopaedia illustration, fine ink lines and flat watercolour fills. Palette: navy, cream, rust red and soft yellow. Composition: the tower centred on a cream background, short labels with thin leader lines on both sides. Text: the labels read "Lamp room", "Spiral stairs", "Keeper bedroom" and "Engine room". Make it a 2:3 frame. Spell the labels exactly as written.

Nano Banana

Generate an image. A technical cutaway illustration of a lighthouse showing its spiral stairs, lamp room, keeper bedroom and engine room. Style: detailed vintage encyclopaedia illustration, fine ink lines and flat watercolour fills. Palette: navy, cream, rust red and soft yellow. Composition: the tower centred on a cream background, short labels with thin leader lines on both sides. Text: the labels read "Lamp room", "Spiral stairs", "Keeper bedroom" and "Engine room". Aspect ratio 2:3. Spell the labels exactly as written.

Midjourney

A technical cutaway illustration of a lighthouse showing its spiral stairs, lamp room, keeper bedroom and engine room, detailed vintage encyclopaedia illustration, fine ink lines and flat watercolour fills, navy, cream, rust red and soft yellow, the tower centred on a cream background, short labels with thin leader lines on both sides, the labels read "Lamp room", "Spiral stairs", "Keeper bedroom" and "Engine room" --ar 2:3 --raw --no text, letters, watermark

Pixel art city at dusk

A palette limit (16 colours) gives true pixel-art restraint.

Any model

An 8-bit pixel art city skyline at dusk with a river, small boats and lit windows. Style: crisp 16-colour pixel art with hard edges, no anti-aliasing and subtle dithering in the sky. Palette: purples, oranges and a few bright window yellows. Composition: a side-scrolling view with the skyline across the middle and a foreground bridge. Format: 16:9. No text, logos or watermarks.

ChatGPT

Create an image. An 8-bit pixel art city skyline at dusk with a river, small boats and lit windows. Style: crisp 16-colour pixel art with hard edges, no anti-aliasing and subtle dithering in the sky. Palette: purples, oranges and a few bright window yellows. Composition: a side-scrolling view with the skyline across the middle and a foreground bridge. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. No text, logos or watermarks.

Gemini

Create an image. An 8-bit pixel art city skyline at dusk with a river, small boats and lit windows. Style: crisp 16-colour pixel art with hard edges, no anti-aliasing and subtle dithering in the sky. Palette: purples, oranges and a few bright window yellows. Composition: a side-scrolling view with the skyline across the middle and a foreground bridge. Make it a 16:9 frame. No text, logos or watermarks.

Nano Banana

Generate an image. An 8-bit pixel art city skyline at dusk with a river, small boats and lit windows. Style: crisp 16-colour pixel art with hard edges, no anti-aliasing and subtle dithering in the sky. Palette: purples, oranges and a few bright window yellows. Composition: a side-scrolling view with the skyline across the middle and a foreground bridge. Aspect ratio 16:9. No text, logos or watermarks.

Midjourney

An 8-bit pixel art city skyline at dusk with a river, small boats and lit windows, crisp 16-colour pixel art with hard edges, no anti-aliasing and subtle dithering in the sky, purples, oranges and a few bright window yellows, a side-scrolling view with the skyline across the middle and a foreground bridge --ar 16:9 --no text, logo, watermark --raw

Picture book bear making soup

Naming the media (gouache, coloured pencil) keeps the style consistent.

Any model

A gentle brown bear in a patched apron stirring a huge pot of soup, woodland friends waiting with bowls. Style: children's picture book illustration, gouache and coloured pencil, soft rounded shapes. Palette: warm autumn palette with muted greens. Setting: a cosy cottage kitchen with hanging herbs and a window onto falling leaves. Composition: the bear on the left, the pot in the centre, animals queuing at the right. Format: 3:2. No text, logos or watermarks.

ChatGPT

Create an image. A gentle brown bear in a patched apron stirring a huge pot of soup, woodland friends waiting with bowls. Style: children's picture book illustration, gouache and coloured pencil, soft rounded shapes. Palette: warm autumn palette with muted greens. Setting: cosy cottage kitchen with hanging herbs and a window onto falling leaves. Composition: the bear on the left, the pot in the centre, animals queuing at the right. Use a landscape frame; if it can't match 3:2 exactly, compose so I can crop to 3:2. No text, logos or watermarks.

Gemini

Create an image. A gentle brown bear in a patched apron stirring a huge pot of soup, woodland friends waiting with bowls. Style: children's picture book illustration, gouache and coloured pencil, soft rounded shapes. Palette: warm autumn palette with muted greens. Setting: cosy cottage kitchen with hanging herbs and a window onto falling leaves. Composition: the bear on the left, the pot in the centre, animals queuing at the right. Make it a 3:2 frame. No text, logos or watermarks.

Nano Banana

Generate an image. A gentle brown bear in a patched apron stirring a huge pot of soup, woodland friends waiting with bowls. Style: children's picture book illustration, gouache and coloured pencil, soft rounded shapes. Palette: warm autumn palette with muted greens. Setting: cosy cottage kitchen with hanging herbs and a window onto falling leaves. Composition: the bear on the left, the pot in the centre, animals queuing at the right. Aspect ratio 3:2. No text, logos or watermarks.

Midjourney

A gentle brown bear in a patched apron stirring a huge pot of soup, woodland friends waiting with bowls, children's picture book illustration, gouache and coloured pencil, soft rounded shapes, warm autumn palette with muted greens, a cosy cottage kitchen with hanging herbs and a window onto falling leaves, the bear on the left, the pot in the centre, animals queuing at the right --ar 3:2 --no text, logo, watermark --raw

Tea stall at dawn

Freezing a pour with a named shutter gives the shot its punch.

Any model

A roadside tea seller pouring chai from a height into small glasses at dawn, steam rising, a few customers waiting. Setting: a city street corner with a tin-roof stall, a kettle on a flame and misty morning air. Light: cool blue dawn with warm flame glow and rim light on the steam. Camera: wide-angle street photography lens, a fast shutter that freezes the stream of tea. Composition: the pour in the centre foreground, customers softly blurred behind. Mood: alive and humble. Format: 3:2. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A roadside tea seller pouring chai from a height into small glasses at dawn, steam rising, a few customers waiting. Setting: city street corner with a tin-roof stall, a kettle on a flame and misty morning air. Lighting: cool blue dawn with warm flame glow and rim light on the steam. Camera: wide-angle street photography lens, a fast shutter that freezes the stream of tea. Composition: the pour in the centre foreground, customers softly blurred behind. Keep the mood alive and humble. Use a landscape frame; if it can't match 3:2 exactly, compose so I can crop to 3:2. No text, logos or watermarks.

Gemini

Create a photorealistic image. A roadside tea seller pouring chai from a height into small glasses at dawn, steam rising, a few customers waiting. Setting: city street corner with a tin-roof stall, a kettle on a flame and misty morning air. Lighting: cool blue dawn with warm flame glow and rim light on the steam. Camera: wide-angle street photography lens, a fast shutter that freezes the stream of tea. Composition: the pour in the centre foreground, customers softly blurred behind. Keep the mood alive and humble. Make it a 3:2 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A roadside tea seller pouring chai from a height into small glasses at dawn, steam rising, a few customers waiting. Setting: city street corner with a tin-roof stall, a kettle on a flame and misty morning air. Lighting: cool blue dawn with warm flame glow and rim light on the steam. Camera: wide-angle street photography lens, a fast shutter that freezes the stream of tea. Composition: the pour in the centre foreground, customers softly blurred behind. Keep the mood alive and humble. Aspect ratio 3:2. No text, logos or watermarks.

Midjourney

A roadside tea seller pouring chai from a height into small glasses at dawn, steam rising, a few customers waiting, a city street corner with a tin-roof stall, a kettle on a flame and misty morning air, cool blue dawn with warm flame glow and rim light on the steam, wide-angle street photography lens, a fast shutter that freezes the stream of tea, the pour in the centre foreground, customers softly blurred behind, alive and humble --ar 3:2 --no text, logo, watermark --raw

Cosy cabin in hand-painted anime style

Background-art wording prevents drifting into character art.

Any model

A small log cabin in a pine forest with lit windows and smoke from the chimney, a path of stepping stones through tall grass. Style: hand-painted anime background art, soft brush texture, detailed clouds. Light: warm evening glow from the windows against a deep teal sky. Composition: a wide establishing view with the cabin on the right third and the path leading in. Format: 16:9. No text, logos or watermarks.

ChatGPT

Create an image. A small log cabin in a pine forest with lit windows and smoke from the chimney, a path of stepping stones through tall grass. Style: hand-painted anime background art, soft brush texture, detailed clouds. Lighting: warm evening glow from the windows against a deep teal sky. Composition: a wide establishing view with the cabin on the right third and the path leading in. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. No text, logos or watermarks.

Gemini

Create an image. A small log cabin in a pine forest with lit windows and smoke from the chimney, a path of stepping stones through tall grass. Style: hand-painted anime background art, soft brush texture, detailed clouds. Lighting: warm evening glow from the windows against a deep teal sky. Composition: a wide establishing view with the cabin on the right third and the path leading in. Make it a 16:9 frame. No text, logos or watermarks.

Nano Banana

Generate an image. A small log cabin in a pine forest with lit windows and smoke from the chimney, a path of stepping stones through tall grass. Style: hand-painted anime background art, soft brush texture, detailed clouds. Lighting: warm evening glow from the windows against a deep teal sky. Composition: a wide establishing view with the cabin on the right third and the path leading in. Aspect ratio 16:9. No text, logos or watermarks.

Midjourney

A small log cabin in a pine forest with lit windows and smoke from the chimney, a path of stepping stones through tall grass, hand-painted anime background art, soft brush texture, detailed clouds, warm evening glow from the windows against a deep teal sky, a wide establishing view with the cabin on the right third and the path leading in --ar 16:9 --no text, logo, watermark --raw

Gold ring on velvet macro

Controlling a single highlight makes metal look expensive.

Any model

A thin gold ring with a small emerald stone resting on deep navy velvet. Light: a single soft window light from the left, a sharp highlight on the metal and a gentle gradient on the velvet. Camera: macro lens, moderate depth of field, sharp on the stone facets. Composition: the ring slightly off centre, a soft fall-off to dark at the edges. Mood: precious and quiet. Format: 1:1. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. A thin gold ring with a small emerald stone resting on deep navy velvet. Lighting: a single soft window light from the left, a sharp highlight on the metal and a gentle gradient on the velvet. Camera: macro lens, moderate depth of field, sharp on the stone facets. Composition: the ring slightly off centre, a soft fall-off to dark at the edges. Keep the mood precious and quiet. Use a square frame; if it can't match 1:1 exactly, compose so I can crop to 1:1. No text, logos or watermarks.

Gemini

Create a photorealistic image. A thin gold ring with a small emerald stone resting on deep navy velvet. Lighting: a single soft window light from the left, a sharp highlight on the metal and a gentle gradient on the velvet. Camera: macro lens, moderate depth of field, sharp on the stone facets. Composition: the ring slightly off centre, a soft fall-off to dark at the edges. Keep the mood precious and quiet. Make it a 1:1 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. A thin gold ring with a small emerald stone resting on deep navy velvet. Lighting: a single soft window light from the left, a sharp highlight on the metal and a gentle gradient on the velvet. Camera: macro lens, moderate depth of field, sharp on the stone facets. Composition: the ring slightly off centre, a soft fall-off to dark at the edges. Keep the mood precious and quiet. Aspect ratio 1:1. No text, logos or watermarks.

Midjourney

A thin gold ring with a small emerald stone resting on deep navy velvet, a single soft window light from the left, a sharp highlight on the metal and a gentle gradient on the velvet, macro lens, moderate depth of field, sharp on the stone facets, the ring slightly off centre, a soft fall-off to dark at the edges, precious and quiet --ar 1:1 --no text, logo, watermark --raw

Aerial view of rice terraces

S curves create flow, and the flooded terraces add the sky reflection.

Any model

Curving rice terraces on a steep hillside, half flooded and mirroring the sky, a few farmers with conical hats. Setting: a green mountain valley with a winding track and a small village. Light: low early sun with long shadows along the terrace edges. Camera: drone view at 100 metres, wide-angle lens, crisp detail. Composition: terrace lines sweeping in S curves across the frame. Mood: serene and orderly. Format: 4:3. No text, logos or watermarks.

ChatGPT

Create a photorealistic image. Curving rice terraces on a steep hillside, half flooded and mirroring the sky, a few farmers with conical hats. Setting: green mountain valley with a winding track and a small village. Lighting: low early sun with long shadows along the terrace edges. Camera: drone view at 100 metres, wide-angle lens, crisp detail. Composition: terrace lines sweeping in S curves across the frame. Keep the mood serene and orderly. Use a landscape frame; if it can't match 4:3 exactly, compose so I can crop to 4:3. No text, logos or watermarks.

Gemini

Create a photorealistic image. Curving rice terraces on a steep hillside, half flooded and mirroring the sky, a few farmers with conical hats. Setting: green mountain valley with a winding track and a small village. Lighting: low early sun with long shadows along the terrace edges. Camera: drone view at 100 metres, wide-angle lens, crisp detail. Composition: terrace lines sweeping in S curves across the frame. Keep the mood serene and orderly. Make it a 4:3 frame. No text, logos or watermarks.

Nano Banana

Generate a photorealistic image. Curving rice terraces on a steep hillside, half flooded and mirroring the sky, a few farmers with conical hats. Setting: green mountain valley with a winding track and a small village. Lighting: low early sun with long shadows along the terrace edges. Camera: drone view at 100 metres, wide-angle lens, crisp detail. Composition: terrace lines sweeping in S curves across the frame. Keep the mood serene and orderly. Aspect ratio 4:3. No text, logos or watermarks.

Midjourney

Curving rice terraces on a steep hillside, half flooded and mirroring the sky, a few farmers with conical hats, a green mountain valley with a winding track and a small village, low early sun with long shadows along the terrace edges, drone view at 100 metres, wide-angle lens, crisp detail, terrace lines sweeping in S curves across the frame, serene and orderly --ar 4:3 --no text, logo, watermark --raw

Coffee brewing infographic poster

Keep text to labels of one to three words, and always give the exact words.

Any model

A clean infographic poster showing four steps of pour-over coffee: grind, rinse, bloom and pour. Text: the title reads "Pour-Over in 4 Steps", with step labels "1 Grind", "2 Rinse", "3 Bloom", "4 Pour". Style: flat vector icons, thin line details, plenty of whitespace. Palette: cream background, espresso brown and one orange accent. Composition: four equal panels in a row with a title above and a short arrow between panels. Format: 16:9. Spell every label exactly as written.

ChatGPT

Create an image. A clean infographic poster showing four steps of pour-over coffee: grind, rinse, bloom and pour. Text: the title reads "Pour-Over in 4 Steps", with step labels "1 Grind", "2 Rinse", "3 Bloom", "4 Pour". Style: flat vector icons, thin line details, plenty of whitespace. Palette: cream background, espresso brown and one orange accent. Composition: four equal panels in a row with a title above and a short arrow between panels. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. Spell every label exactly as written.

Gemini

Create an image. A clean infographic poster showing four steps of pour-over coffee: grind, rinse, bloom and pour. Text: the title reads "Pour-Over in 4 Steps", with step labels "1 Grind", "2 Rinse", "3 Bloom", "4 Pour". Style: flat vector icons, thin line details, plenty of whitespace. Palette: cream background, espresso brown and one orange accent. Composition: four equal panels in a row with a title above and a short arrow between panels. Make it a 16:9 frame. Spell every label exactly as written.

Nano Banana

Generate an image. A clean infographic poster showing four steps of pour-over coffee: grind, rinse, bloom and pour. Text: the title reads "Pour-Over in 4 Steps", with step labels "1 Grind", "2 Rinse", "3 Bloom", "4 Pour". Style: flat vector icons, thin line details, plenty of whitespace. Palette: cream background, espresso brown and one orange accent. Composition: four equal panels in a row with a title above and a short arrow between panels. Aspect ratio 16:9. Spell every label exactly as written.

Midjourney

A clean infographic poster showing four steps of pour-over coffee: grind, rinse, bloom and pour, the title reads "Pour-Over in 4 Steps", with step labels "1 Grind", "2 Rinse", "3 Bloom", "4 Pour", flat vector icons, thin line details, plenty of whitespace, cream background, espresso brown and one orange accent, four equal panels in a row with a title above and a short arrow between panels --ar 16:9 --raw

Surreal whale above a wheat field

A small familiar object gives surreal scenes a sense of scale.

Any model

A huge blue whale floating silently above a golden wheat field at dusk, its shadow rolling across the crops. Setting: a flat farmland horizon with a lone farmhouse and a dirt road. Light: low orange sunset behind, rim light on the whale, long shadows. Camera: wide-angle lens, deep focus, painterly cinematic grade. Composition: the whale filling the upper centre, the farmhouse small in the lower left for scale. Mood: dreamlike and calm. Format: 16:9. No text, logos or watermarks.

ChatGPT

Create an image. A huge blue whale floating silently above a golden wheat field at dusk, its shadow rolling across the crops. Setting: flat farmland horizon with a lone farmhouse and a dirt road. Lighting: low orange sunset behind, rim light on the whale, long shadows. Camera: wide-angle lens, deep focus, painterly cinematic grade. Composition: the whale filling the upper centre, the farmhouse small in the lower left for scale. Keep the mood dreamlike and calm. Use a wide landscape frame; if it can't match 16:9 exactly, compose so I can crop to 16:9. No text, logos or watermarks.

Gemini

Create an image. A huge blue whale floating silently above a golden wheat field at dusk, its shadow rolling across the crops. Setting: flat farmland horizon with a lone farmhouse and a dirt road. Lighting: low orange sunset behind, rim light on the whale, long shadows. Camera: wide-angle lens, deep focus, painterly cinematic grade. Composition: the whale filling the upper centre, the farmhouse small in the lower left for scale. Keep the mood dreamlike and calm. Make it a 16:9 frame. No text, logos or watermarks.

Nano Banana

Generate an image. A huge blue whale floating silently above a golden wheat field at dusk, its shadow rolling across the crops. Setting: flat farmland horizon with a lone farmhouse and a dirt road. Lighting: low orange sunset behind, rim light on the whale, long shadows. Camera: wide-angle lens, deep focus, painterly cinematic grade. Composition: the whale filling the upper centre, the farmhouse small in the lower left for scale. Keep the mood dreamlike and calm. Aspect ratio 16:9. No text, logos or watermarks.

Midjourney

A huge blue whale floating silently above a golden wheat field at dusk, its shadow rolling across the crops, a flat farmland horizon with a lone farmhouse and a dirt road, low orange sunset behind, rim light on the whale, long shadows, wide-angle lens, deep focus, painterly cinematic grade, the whale filling the upper centre, the farmhouse small in the lower left for scale, dreamlike and calm --ar 16:9 --no text, logo, watermark --raw

Edit: recolour a sofa in my room photo

State the one change, then list what to keep. This stops stray edits.

Attach your photo first. “Edit: recolour a sofa in my room photo” edits a photo you upload, so it won't work without one.

ChatGPT

If no photo is attached or you can't edit images, say so in one sentence and stop. Edit my uploaded photo of a living room. Change: only the sofa: make it a dark emerald green velvet with the same shape, cushions and shadows. Keep: the walls, floor, rug, windows, lighting and every other object exactly as they are. Check: the velvet should catch the light from the window in the same direction as the original. Do not change anything else. No text, letters, numbers, logos or watermarks.

Gemini

If no photo is attached or you can't edit images, say so in one sentence and stop. Edit my uploaded photo of a living room. Change: only the sofa: make it a dark emerald green velvet with the same shape, cushions and shadows. Keep: the walls, floor, rug, windows, lighting and every other object exactly as they are. Check: the velvet should catch the light from the window in the same direction as the original. Do not change anything else. No text, letters, numbers, logos or watermarks.

Nano Banana

If no photo is attached or you can't edit images, say so in one sentence and stop. Edit my uploaded photo of a living room. Change: only the sofa: make it a dark emerald green velvet with the same shape, cushions and shadows. Keep: the walls, floor, rug, windows, lighting and every other object exactly as they are. Check: the velvet should catch the light from the window in the same direction as the original. Do not change anything else. No text, letters, numbers, logos or watermarks.

How to use
  1. Copy it with the button above.
  2. Fill the blanks in {{double braces}} with your own details.
  3. Paste it under your photo in your image AI and generate.
  4. Then ask for one change at a time, not a new prompt.

Edit: remove passers-by from a travel photo

Asking to rebuild what is behind them reduces smudging.

ChatGPT

Edit my uploaded travel photo of a famous square. Change: remove all the people in the background and rebuild the cobblestones, shopfronts and shadows where they stood. Keep: the main subject in the foreground, the sky, the lighting and the colour exactly as they are. Check: no blur patches, smudges or repeated textures where people were removed. Do not change anything else. No text, letters, numbers, logos or watermarks.

Gemini

Edit my uploaded travel photo of a famous square. Change: remove all the people in the background and rebuild the cobblestones, shopfronts and shadows where they stood. Keep: the main subject in the foreground, the sky, the lighting and the colour exactly as they are. Check: no blur patches, smudges or repeated textures where people were removed. Do not change anything else. No text, letters, numbers, logos or watermarks.

Nano Banana

Edit my uploaded travel photo of a famous square. Change: remove all the people in the background and rebuild the cobblestones, shopfronts and shadows where they stood. Keep: the main subject in the foreground, the sky, the lighting and the colour exactly as they are. Check: no blur patches, smudges or repeated textures where people were removed. Do not change anything else. No text, letters, numbers, logos or watermarks.

How should you write a prompt for Gemini image generation?

A prompt for Gemini image generation should describe the scene in full sentences: subject, setting, light, camera or art style, composition and aspect ratio. Gemini follows sentences better than keyword lists, so write the scene as two full sentences. Paste a prompt, then change one detail at a time.

Before you start

**There are 25 prompt examples for image generation here**, including a ceramic mug product shot, a fjord at blue hour, an isometric bakery cutaway and a vintage travel poster with text. They run in groups: products and food, places and interiors, nature, illustration styles, posters and infographics, and edits. Compare the keyword list "mug, oak table, steam, window light, 85mm, 4:5" with its sentence rewrite: "A matte grey mug sits on an oak table with steam rising, lit by soft window light, shot on an 85mm lens, framed 4:5." The sentence fills every gap, and that's why the cards read as sentences. Of the 25, 23 generate a new image and 2 edit a photo you upload. None of them use a face, so you don't need a reference photo. For portraits that keep your own face, jump to the pictures of yourself prompts instead.

Prompt examples for image generation in Gemini

There are 25 prompt examples for image generation in Gemini here, including a ceramic mug product shot, a fjord at blue hour, an isometric bakery cutaway and a vintage travel poster with text.

They're sorted by job: products and food, places and interiors, nature, illustration styles, posters and infographics, then 2 edit cards that change a photo you've uploaded (a sofa recolour and a passer-by removal). Pick the closest card, swap the subject words and keep the light and lens lines as they are.

Can Gemini put text in a generated image?

Yes, Gemini can put short text in a generated image, such as a 1 to 5 word poster title in quotation marks, but check each letter before you use the image. Put the exact words in quotes and say where they go. The vintage travel poster and the coffee brewing infographic cards show the pattern. If one word comes out wrong, ask Gemini to fix only that word instead of regenerating the whole picture.

Gemini edits photos too. Upload a photo, say what to change, then list what must stay the same. The two edit cards recolour a sofa in a room photo and remove passers-by from a travel photo. For a scratched or faded print, the old photo restoration prompts are a better fit.

Text isn't perfect every time, so it's worth a quick check. You'll spot a typo in seconds, and you won't need to start over: ask for the single fix. Don't paste long sentences either, since they're the ones that go wrong.

Why does Gemini ignore parts of an image prompt?

Gemini ignores parts of an image prompt when the prompt packs in too many subjects or conflicting styles; keep one subject, one style and put the ratio last. Four slips cause most misses:

  • Too many subjects in one scene, so the last ones get dropped.
  • A keyword list in place of sentences.
  • Two styles that fight, such as flat vector next to a photorealistic lens.
  • An aspect ratio buried mid-prompt, where it's easy to lose.

Fix them one at a time and rerun. Cutting a prompt back to a single subject usually brings the missing part back.

Don't worry if the first try isn't right. It's normal, it doesn't mean the prompt is broken, and you'll usually fix it in a rerun or two. Here's a good habit: change one thing, look, then change the next.

ChatGPT and Midjourney versions of the Gemini image prompts

The Gemini image prompts run in ChatGPT as written, while Midjourney needs them cut to short phrases with --ar for the ratio. ChatGPT takes the same sentences, but it offers fixed frame sizes, so its version asks for the nearest shape. Midjourney prefers a comma-separated description that ends with the --ar flag. Whichever model you pick, check any quoted text letter by letter.

As of 10 Nov 2026, Google's Gemini Application Programming Interface (API) model page lists the current Nano Banana image model as gemini-nano-banana-2.1, which can output 1K, 2K or 4K. Model names change, so check ai.google.dev. The Nano Banana prompt builder helps you fill in each part.

For a scene with your own family in it instead of an invented one, switch to an AI family photo prompt. For two adults and two children in one frame, each card of the family picture of 4 prompts page sets the height order. Gemini prompts for writing, study and work sit with the image ones under prompts for Gemini. The realistic photo prompts cover camera technique, and the Gemini photo prompts hub lists every photo pack.

There's no hand-converting to do. Each card has the ChatGPT and Midjourney versions ready, so there's nothing to convert, and they're one click away.

That's all it takes, and it's the same method every time.

Gemini prompt for image generation FAQ

Still stuck? Build it in Prompt Studio. It'll score each part as you go.

What is the best Gemini prompt for image generation?

The best Gemini prompt for image generation is 2 or 3 full sentences that name the subject, the light and the style, ending with the ratio, like the fjord at blue hour card. It's specific without being long, so skip the padding and swap any detail to make it yours.

Is Gemini image generation free?

Yes, Gemini image generation is free in the Gemini app within a daily limit, and paid plans raise that limit. Google sets the numbers per plan, so check what the app shows for your account. The prompts here work the same on every plan; you'll just get fewer tries on a free one.

Can Gemini edit a photo I upload?

Yes, Gemini can edit a photo you upload: the 2 edit cards here recolour a sofa and remove passers-by while keeping the rest of the photo the same. Say what to change, then list what to keep, and change one detail per run.

How long should a Gemini image prompt be?

A Gemini image prompt should be 2 to 4 sentences long, since one keyword line leaves gaps and a long paragraph full of styles makes Gemini drop parts. Aim for one subject, one style and a ratio at the end.

Build your own image prompt

Answer a few questions. The prompt's scored as it's built. It's free, and nothing leaves your browser.