Rune

Writing good prompts at scale is one of the biggest time sinks in AI stock production. Rune solves this with a matrix system — you define the key variables of your images (subject, style, mood, setting) and Rune generates every possible combination automatically. What would take hours to write manually takes seconds.

Define your inputs as separate axes. For example:

· Subject: "woman", "man", "couple"

· Style: "minimalist", "cinematic", "editorial"

· Mood: "confident", "calm", "energetic"

· Setting: "modern office", "outdoor cafe"

Rune combines every variation — 3 × 3 × 3 × 2 = 54 unique prompts in one click. Scale up your axes and you can generate thousands. The more specific your inputs, the more consistent and usable the results.

Have a reference image whose visual style you love — from a shoot, a competitor's portfolio, or your own past work? Drop it into the Style Extractor. Rune analyzes the composition, lighting, color grading, and mood, then turns it into a reusable style component you can inject into any batch.

This is the fastest way to maintain a consistent visual identity across a large batch of AI-generated images.

Every generated prompt lands in a review queue before anything gets sent to image generation. Go through them — approve the ones that make sense, edit the ones that are close, and delete the ones that aren't relevant. Nothing moves forward without your sign-off.

You can also edit prompts directly in the queue to add specifics, negative descriptions, or technical parameters before approving.

Specificity wins. "Senior businessman in a modern glass office, confident expression, natural window light" will produce far more consistent and useful results than "person working".

Avoid vague moods. "Happy" is weak. "Calm, focused, quietly productive" gives the model more to work with.

Include photographic intent. Adding terms like "stock photo", "professional photography", "editorial style" steers the output toward images that actually sell.