An image prompt is a text description that tells an AI image generator what to create. It usually includes:
- Subject (what is in the image)
- Style (e.g. cinematic, watercolor, 3D, cyberpunk)
- Lighting (soft daylight, neon rim light, studio softbox)
- Composition (close-up, wide shot, bird’s-eye view, centered)
- Details and constraints (camera lens, aspect ratio, mood, background)
AI models are sensitive to wording. Small changes can strongly affect:
- Color palette and atmosphere
- The “camera” look (lens, depth of field)
- Level of realism vs illustration
- Consistency across multiple generations
Use this as a reliable starting point:
- Subject: who/what is the main focus?
- Scene: where is it, what is happening?
- Style: art movement, medium, visual reference
- Lighting: time of day, key light, mood
- Camera: focal length, depth of field, shot type
- Quality: sharp, detailed, clean, cinematic, etc.
Example:
portrait of a traveler in a rainy neon city, cyberpunk style, cinematic lighting, shallow depth of field, 85mm lens, ultra detailed, high contrast
If you have a reference image you like, Image to Prompt can help you extract:
- Style and aesthetic cues
- Lighting and color notes
- Composition and perspective hints
- A prompt you can copy, paste, and iterate on