Luma Labs Releases 25 Gemini Photo Editing Prompts For Image Prep

Luma Labs published a practical prompt guide featuring 25 specialized image editing instructions for Google Gemini. The collection outlines concrete text prompts designed to handle portrait retouching, commercial product cleanup, archival restoration, and targeted composite adjustments. By standardizing multi-modal editing inputs, the guide addresses visual consistency issues that frequently disrupt AI pre-production workflows.

What's new

Luma Labs structured the guide into four focused categories: portrait retouching, product presentation, photo restoration, and creative adjustments. Rather than relying on generic descriptive phrasing, each example provides exact prompt formulas tailored to Gemini's multi-modal understanding. The techniques cover localized lighting fixes, background isolation, noise reduction on archival scans, and precise object swap commands.

The released prompt workflows emphasize non-destructive adjustments. By specifying camera parameters, focal length, light direction, and subject boundaries in the prompt string, editors can modify specific image elements without corrupting neighboring visual details. Luma Labs included side-by-side visual comparisons for all 25 examples to demonstrate expected outputs across varying image resolutions.

How it fits your workflow

For filmmakers, storyboard artists, and VFX supervisors using image-to-video engines like Luma Dream Machine, high-quality reference stills are critical. Feeding inconsistent or artifact-heavy source images into motion generators produces flickers, warp distortion, and broken character tracking. Using Gemini to clean up portraits or stage assets prior to video generation creates clean base plates that hold up better under motion synthesis.

This workflow serves as a lightweight alternative to manual Photoshop retouching or dedicated upscaling tools like Magnific AI and Krea AI. While Midjourney v6 and Flux 1.1 Pro require re-rolling entire image compositions for minor changes, Gemini's instruction-following capability allows editors to request targeted localized edits on existing photography. Pre-visualization teams can rapidly fix lighting mismatches or remove background distractions across an entire deck of concept images before animating them in Luma Dream Machine or Runway Gen-3 Alpha.

What it costs / how to try it

The 25 prompt examples are available for free on the official Luma Labs website as of mid-2024. Editors can copy the prompt strings directly into the Gemini interface or integrate them into custom API scripts for batch photo processing.