Skip to main content
Eight Labs·Prompts

ChatGPT Image Prompts
That Actually Work

40+ prompts across seven capabilities — precision, multilingual text, photorealism, art styles, aspect ratios, real-world intelligence, and multi-image workflows. Copy and use directly.

Apr 25, 2026·Seven categories·40+ prompts·45 images
01

Greater precision and control

The model holds fine detail across texture, handwriting, spatial layout, and object relationships. These prompts push it on specificity — and it delivers.

Prompt

a photorealistic, taken by phone photo of a handwritten essay in pencil, bold but elegant handwriting, but messy and somewhat uneven, on an 8.5x11 piece of lined paper, about the history of baseball in toronto. make sure there is variance in the writing in a very human way. give it a slight coffee stain on the top right corner

Prompt

A high-fashion editorial photograph of a model wearing an elegant, structured outfit in the middle of a rice paddy field in rural Japan. The contrast between high fashion and natural agricultural setting. Golden hour lighting, shot on medium format film.

Prompt

Create a visual polyglot — a single image that contains handwritten text, typeset text, a diagram, and a photograph-within-a-photograph, all arranged as if placed on a wooden desk.

02

Stronger across languages

Non-Latin scripts used to be where image generation fell apart. Not anymore. These prompts produce coherent, design-integrated text in Japanese, Korean, Chinese, Hindi, and Bengali.

Prompt

Generate professional multilingual poster about typography. The poster is supposed to be an artwork celebrating languages around the world. Japanese editorial style. 4:5 portrait aspect ratio

Prompt

Design a Japanese vintage travel advertisement poster for a mountain onsen resort. Include Japanese calligraphy as the headline, a woodblock print aesthetic, muted earth tones. Portrait format.

Prompt

Create a traditional Chinese ink-wash painting style poster advertising a tea ceremony. Include elegant Chinese characters as the title. Horizontal format, wide aspect ratio.

03

Stylistic sophistication and realism

Photorealism, cinematic stills, film grain, lens characteristics — the model understands what makes a photograph look real and applies it deliberately.

Prompt

35mm photograph of a book of 1970s NYC candid street photographs — the kind you'd find in a used bookshop. Show the book open to a spread, slightly worn pages, natural ambient light from a window.

Prompt

Candid documentary-style photographs of people at work in a 1980s university computer lab. Fluorescent lighting, CRT monitors, period clothing. Shot on film, slightly underexposed.

Prompt

A cinematic still from a sci-fi film where alien beings observe a human settlement from a distance. Anamorphic lens flare, wide format, overcast light, high detail.

Prompt

High-fashion editorial — an avant-garde runway look photographed against a brutalist concrete structure at dusk. Shot on a digital medium format camera with a 85mm lens.

Prompt

Photorealistic studio portrait with dramatic chiaroscuro lighting. Subject looking off-camera. Shot on film, slight grain, warm tone.

04

Styles

The model doesn't approximate styles — it inhabits them. Manga, indie comics, mid-century French design, shounen action sequences. These prompts unlock the full range.

Prompt

A page of a comic book in the style of modern indie comic — clean linework, limited palette, expressive character design, panels with white gutters. The scene: two people having a tense conversation in a diner at night.

Prompt

A vibrant Miami-style comic book cover from the 1980s — neon colors, bold outlines, action pose, palm trees. Retro printing texture with slight halftone dots.

Prompt

A full-page shounen manga spread — explosive action, speed lines, dramatic perspective, Japanese sound effects, panel layout that conveys kinetic energy.

Prompt

Pixel art RPG environment — a small medieval town at dusk, 16-bit style, warm lantern glow, cobblestone streets, NPCs mid-animation. 4:3 format.

05

Flexible aspect ratios

Landscape panoramas, tall editorial columns, square social graphics, narrow bookmarks. The model composes intentionally for the format, not just crops.

Prompt

a really tall and narrow photo where there's a winding path that goes through children storybook-like characters and phrases like "not yet" until it arrives at a satisfying conclusion at the bottom. it should be on a white background and the path should be a really thin winding solid black line

Prompt

A wide panoramic photograph of lavender fields in Provence, France at golden hour. Ultra-wide 3:1 format, natural lens distortion, impressionist color palette, shot on medium format film.

Prompt

A manga-style technical disassembly diagram of a vintage mechanical watch — exploded view, labelled components in Japanese, clean white background, blueprint aesthetic.

Prompt

A traditional Chinese landscape painting in the Song Dynasty style — mountains receding into mist, a lone pine tree, a small pavilion, minimal ink strokes, generous negative space. Vertical scroll format.

06

Real-world intelligence

This is where image generation becomes a research and analysis tool. The model synthesises information, constructs visual arguments, and outputs something immediately usable.

Prompt

Using this portrait, create a diagram-first personal color analysis. Show which clothing colors suit the subject through visual comparison. Keep text minimal and avoid paragraphs.

Prompt

Create an infographic showing global wheat production trends from 2000–2024. Include regional breakdowns, key events that affected supply (drought years, conflict), and projections. Data-journalism style, clean typography, muted earth-tone palette.

Prompt

Create a visual proof of Cantor's diagonal argument showing that real numbers are uncountable. Include a partially filled infinite table of decimal expansions, highlight the diagonal, and show the constructed number. Academic poster style, clean layout.

07

A visual thought partner

With thinking mode, the model generates multiple coherent images in one run. Brand systems, comic sequences, UI directions — whole projects, not single shots.

Prompt

a 35mm film photograph of a blackboard in a classroom, and on the blackboard is a visual proof of sum of odd numbers is a square. there is a title "Thinking Mode On" before the math proof. Aspect Ratio: 4:5 portrait.

Prompt

Generate a complete brand identity system for a specialty coffee roaster called 'Meridian'. Include: logo mark, wordmark, color palette swatch card, and a sample bag label. All on a clean white background. Deliver as a single image.

Prompt

Create a 4-page manga sequence showing a robot becoming self-aware — wordless, emotion conveyed through posture and environment, consistent character design across all pages.

Prompt

Academic research poster on the relationship between sleep cycles and creative problem-solving. Include an abstract, methodology diagram, key findings, and visual data representation. A4 portrait format, clean academic style.

Eight Labs · TheAIHow

Building with AI image generation?

We cover AI tools, APIs, and production patterns for builders who ship. New tutorials weekly.

Watch on TheAIHow