A Practical Guide to Writing AI Image Prompts

Good image prompts read like a brief to a photographer or illustrator. Here's a structure that works across most tools.

1. Subject: who or what is in the picture.
2. Setting: where and when.
3. Style: photograph, watercolour, flat vector, 3D render, oil painting.
4. Composition: close-up, wide shot, overhead, centred, rule of thirds.
5. Lighting: golden hour, soft window light, studio, neon.
6. Colour and mood: warm, muted, bright, moody.
7. Camera details: lens or depth of field, if you want photographic realism.

A ceramic coffee mug on a wooden table by a window, morning light, shallow depth of field, warm tones, product photography, clean uncluttered background.

• Be specific about the main subject. Vague prompts give generic results.
• Put text in quotation marks and say where it goes.
• Use reference images for character, brand or style consistency.
• Iterate: change one thing at a time.
• Edit instead of regenerating when the tool supports conversational edits.
• State aspect ratio for the platform you need (square, portrait, wide).

• Hands and fine details can still look wrong. Check closely.
• Brand logos and real people: avoid them unless you have the rights and consent.
• Mismatch with real products: if it's for selling, the image must match the item.

Check your tool's commercial terms, avoid imitating a living artist's style for commercial use, and label AI images where platforms or rules require it.

Describe subject, setting, style, light and composition, then refine step by step. A clear brief beats a long prompt.