A lighting vocabulary for better AI images
Beginners describe the subject; experienced prompt-writers describe the light. The way a scene is lit decides its mood, depth and realism more than almost anything else, and AI models respond strongly to named lighting setups borrowed from photography and cinematography. This library collects the most useful ones — searchable by keyword and filterable by mood — with copy-ready prompt phrasing.
How lighting terms work
Image models learned from captioned photographs and film stills, so phrases like “golden hour,” “split lighting” or “volumetric god rays” map onto real visual patterns: shadow direction, contrast ratio, color temperature and atmosphere. Choosing a lighting term effectively art-directs the render:
- Warm vs cool sets the emotional temperature (golden hour vs blue hour).
- Soft vs hard controls shadow edges (studio softbox vs harsh noon sun).
- High key vs low key sets the overall brightness and contrast feel.
- Directional setups like Rembrandt, rim and split lighting sculpt the subject’s form.
Tips for using lighting prompts
- Lead with light when mood matters. If the feeling of the image is the point, put the lighting term near the front of the prompt.
- One setup at a time. Pick a single primary lighting style and, at most, one atmospheric modifier like fog or haze.
- Pair with camera terms from the related Camera & Lens picker for photographic realism, or with art-style terms for a stylised look.
- Match light to time of day. Golden hour, blue hour, moonlight and candlelight all imply specific times — keep them consistent with the rest of your scene.