--- title: "Gemini Photo Editing Prompts: 25 Retouching Examples" description: "Gemini photo editing prompts for portraits, products, restoration, and creative edits. Explore 25 retouching examples with practical prompts for consistent AI photo workflows." canonical: "https://lumalabs.ai/news/gemini-photo-editing-prompts" source: "https://lumalabs.ai/news/gemini-photo-editing-prompts.md" --- # Gemini Photo Editing Prompts: 25 Retouching Examples _By Luma team · September 29, 2026_ A single headshot can live across LinkedIn profiles, press kits, speaking engagements, and internal directories. A product photo might need white backgrounds for e-commerce, lifestyle contexts for social, and isolated assets for print. The brief rarely changes, but the deliverables multiply. Gemini AI photo editing prompts give you a way to handle these variations without rebuilding every image from scratch. The difference between useful prompts and wasted time comes down to structure. You need to know exactly what to ask for, what to preserve, and how to phrase the edit so results match your brief. This guide covers 25 retouching prompts organized by the work they actually accomplish, from portrait touch-ups to product polish to vintage restoration. ## **Key Takeaways** - **Effective Gemini prompts** follow a consistent structure: subject, action, preservation rules, and style parameters - Phrases like** "keep face identical"** prevent unwanted changes to facial features during edits - **Portrait retouching prompts** work best when you specify texture preservation to avoid over-smoothing - **Product photography prompt**s require explicit background and lighting instructions for commercial use - **Using specific terms** like "85mm lens" or "soft studio light" produces more predictable results - **Restoration prompts** need separate instructions for color correction versus structural repair [Try These Prompts Now](https://auth.lumalabs.ai/sign-up) ## **Understanding AI Photo Editing for Creative Workflows** AI photo editing has moved past novelty into daily production work. You can now use these tools to handle repetitive retouching tasks, freeing time for the decisions that actually shape campaigns. The shift matters because volume has increased while timelines have compressed. A brand photoshoot that once delivered 20 selects now needs to generate 200 variants across channels. Your original creative direction stays constant, but the execution scales through intelligent editing. Gemini's approach to photo editing works through natural language prompts. You describe what you want changed, what should stay the same, and the visual parameters that matter. The model processes these instructions against the uploaded image and returns an edited version. ### **The Prompt Formula That Works** Every effective Gemini photo editing prompt contains four elements: subject identification, action specification, preservation rules, and style parameters. Miss any element and results become inconsistent. Subject identification tells the model what to focus on. Action specification describes the change. Preservation rules protect everything that should not change. Style parameters set the visual quality standards. For teams working on campaigns with approved creative, the preservation rules matter most. You want to change one element while keeping everything else intact, a workflow that [Layers handles natively](https://lumalabs.ai/news/introducing-layers) when working with images generated or uploaded into Luma's platform. ## **Portrait Retouching Prompts** Portrait work requires the most careful balance between improvement and authenticity. Over-editing destroys the natural qualities that make portraits work. Under-editing leaves distracting elements that pull focus. ### **1. Natural Skin Retouching** **Prompt Structure** - Subject: The person's skin - Action: Smooth minor texture - Preservation: Keep pores, natural highlights, face shape, skin color, eye and hair detail - Style: Natural, not over-processed **Example Prompt** _"Perform natural skin retouching on the subject, smoothing only minor texture while keeping pores and natural highlights visible. Remove tiny blemishes without touching face shape or altering skin color. Keep eye and hair detail sharp."_ **Why It Works** This prompt works because it specifies what to preserve alongside what to change. The explicit texture preservation prevents the plastic, over-smoothed look that ruins portrait retouching. You get cleaner skin that still looks like real skin. ### **2. Portrait Lighting Enhancement** **Prompt Structure** - Subject: Overall lighting on the portrait - Action: Enhance to mimic soft studio light from a specific direction - Preservation: Natural skin tones, realistic highlights - Style: Soft studio quality with increased midtone contrast **Example Prompt:**_ "Enhance portrait lighting to mimic soft studio light from camera left. Lift shadows slightly, add midtone contrast, keep highlights natural, and maintain realistic skin tone."_ **Why It Works** Lighting prompts benefit from specific vocabulary. Specifying light direction (camera left) and quality (soft studio) gives Gemini clear parameters rather than subjective interpretation. You get consistent results across multiple images. ### **3. Blemish and Imperfection Fix** **Prompt Structure** - Subject: Small blemishes, stray hairs, sensor dust - Action: Remove using texture-matched healing - Preservation: Approved facial features - Style: Natural healing that matches surrounding area **Example Prompt**_: "Target small blemishes, stray hairs, and sensor dust only, using healing that matches natural texture. Apply changes only to areas under 5mm in the final crop."_ **Why It Works** Size limits prevent over-correction. Without them, the model might interpret "imperfections" too broadly and alter features your client approved. The 5mm specification keeps edits focused on actual flaws. ### **4. Realistic Makeup Enhancement** **Prompt Structure** - Subject: Existing makeup (eyeshadow, eyeliner, lips) - Action: Enhance subtly with percentage-based increases - Preservation: Natural skin texture, avoid artificial shine - Style: Enhanced but believable makeup **Example Prompt:**_ "Enhance existing makeup subtly by deepening eyeshadow, sharpening eyeliner edges, and increasing lip color by about 10 percent. Avoid artificial shine and preserve natural skin texture."_ **Why It Works** Percentage controls give predictable results. When your creative director asks for "a bit more," you can increment systematically (10%, 20%, 30%) rather than guessing. This makes revisions faster and more consistent. ### **5. Eye Enhancement** **Prompt Structure** - Subject: The subject's eyes - Action: Brighten, clarify, add subtle catchlight (the light reflection in the eye) - Preservation: Natural eye color, pupil size - Style: Clear, engaging eyes without artificial enhancement **Example Prompt:** _"Brighten and clarify the subject's eyes by adding subtle catchlight, increasing iris detail, and lifting shadows around the eye socket. Maintain natural eye color and pupil size."_ **Why It Works** Eye edits require explicit pupil and color preservation. Without these rules, results can drift toward unnatural enhancement that looks fake. The catchlight addition (a small light reflection) makes eyes appear more engaged without changing their fundamental appearance. ### **6. Teeth Whitening** **Prompt Structure** - Subject: Teeth - Action: Reduce yellow tones and even out color - Preservation: Natural variation, tooth shape, gum line - Style: Natural whitening, not bleached **Example Prompt:** _"Whiten teeth naturally by reducing yellow tones and evening out color, keeping natural variation and avoiding pure white. Preserve tooth shape and gum line."_ **Why It Works** The "avoiding pure white" instruction prevents the overly bleached look that immediately reads as edited. Natural teeth have subtle color variation. This prompt whitens while maintaining that believability. ### **7. Wrinkle Softening** **Prompt Structure** - Subject: Deep wrinkles and lines - Action: Reduce shadow depth - Preservation: Skin texture, expression lines, natural facial structure - Style: Softened but character-preserving **Example Prompt:** _"Soften deep wrinkles and lines by reducing shadow depth without eliminating skin texture. Keep expression lines and natural facial structure visible."_ **Why It Works** Wrinkle work fails when it removes character. This prompt preserves the face while reducing distracting depth. You maintain the person's natural age and personality while creating a more polished look. ### **8. Hair Color Adjustment** **Prompt Structure** - Subject: Hair color - Action: Shift toward a specific color - Preservation: Individual hair strands, highlight and shadow variation - Style: Natural color with depth, not flat **Example Prompt:** _"Shift hair color toward warm auburn while maintaining highlight and shadow variation. Preserve individual hair strands and avoid flat color application."_ **Why It Works** Hair edits need explicit strand preservation. Without it, hair becomes a solid color block that looks painted on. This prompt changes the overall color while keeping the natural variation that makes hair look real. ## **Photo Cleanup and Correction Prompts** Cleanup work removes distractions so the subject commands attention. These prompts handle the technical problems that happen during shoots. ### **9. Background Distraction Removal** **Prompt Structure** - Subject: Distracting background elements - Action: Remove and fill with matching texture - Preservation: Subject edges, original depth, existing blur - Style: Seamless integration **Example Prompt:** _"Remove distracting background elements while preserving original depth and bokeh. Patch gaps with matching texture and avoid changes to subject edges."_ **Why It Works** The texture-matching instruction ensures seamless integration rather than obvious patches. "Bokeh" refers to the quality of the out-of-focus areas. By preserving it, you maintain the photograph's original aesthetic and depth. ### **10. Outdoor Tone Correction** **Prompt Structure** - Subject: Color cast on skin from environmental reflections - Action: Correct green/blue cast, restore natural skin tones - Preservation: Accurate environmental colors in background - Style: Natural, balanced color **Example Prompt:** _"Correct green color cast from foliage reflections on skin, restoring natural skin tones while maintaining accurate environmental colors in the background."_ **Why It Works** Outdoor portraits often pick up color casts from surroundings (green from grass and trees, blue from sky). This prompt targets the problem on skin without over-correcting the scene, which would make the environment look unnatural. ### **11. Stray Hair and Dust Removal** **Prompt Structure** - Subject: Flyaway hairs outside main hair silhouette, sensor dust - Action: Remove - Preservation: Natural hair outline, styled hair - Style: Clean but natural hair edges **Example Prompt:** _"Remove flyaway hairs outside the main hair silhouette and sensor dust spots. Preserve the natural hair outline and avoid altering styled hair."_ **Why It Works** The distinction between flyaways and styled hair prevents unwanted changes to intentional hairstyling. You clean up messy edges without losing the hairstyle your subject or stylist created. ### **12. Fix Blurry Photo** **Prompt Structure** - Subject: Slightly soft focus on face - Action: Sharpen - Preservation: Natural depth of field, avoid edge halos - Style: Sharp but natural **Example Prompt:** _"Sharpen slightly soft focus on the subject's face while maintaining natural depth of field falloff. Increase clarity without adding artificial edge halos."_ **Why It Works:** Sharpening prompts need halo prevention. Over-sharpening creates bright edge artifacts (halos) that destroy the edit. This prompt brings back sharpness while avoiding that telltale over-processed look. ### **13. Remove Unwanted Person** **Prompt Structure** - Subject: Specific person in background - Action: Remove and fill with environment - Preservation: Main subject completely untouched - Style: Seamless environmental fill **Example Prompt:**_ "Remove the background person on the left side of frame, filling the space with matching environment texture. Preserve the main subject completely."_ **Why It Works** For [precision editing tasks](https://lumalabs.ai/news/ai-photo-editing-prompts) like this, explicit location (left side) and fill instructions prevent ambiguity. You get exactly the removal you want without affecting your main subject. ## **Product and E-Commerce Prompts** Product photography serves commercial goals. These prompts prepare images for catalog use, social media, and marketing materials. ### **14. Product Polish and Sharpen** **Prompt Structure** - Subject: Product surfaces - Action: Remove dust and minor scratches, enhance edge definition - Preservation: Accurate product color - Style: Clean, sharp product presentation **Example Prompt:** _"Polish product surfaces to remove dust and minor scratches. Enhance edge definition and increase clarity on texture details. Maintain accurate product color."_ **Why It Works** Product edits prioritize accuracy. Color shifts or texture changes misrepresent the product and create customer service problems. This prompt cleans up production flaws without changing what the product actually looks like. ### **15. White Background Product Shot** **Prompt Structure** - Subject: Background - Action: Replace with pure white - Preservation: Natural product shadows, edge definition, accurate color - Style: E-commerce standard white background **Example Prompt:** _"Replace the current background with pure white while preserving natural product shadows and edge definition. Maintain accurate product colors throughout."_ **Why It Works:** The shadow preservation ensures your product doesn't look like it's floating. E-commerce platforms often require pure white backgrounds, and this prompt delivers that standard while keeping your product looking naturally photographed. ### **16. Lifestyle Context Shot** **Prompt Structure** - Subject: Product placement - Action: Place in specific environment with appropriate lighting - Preservation: Product proportions, accurate product appearance - Style: Natural product placement with proper shadows **Example Prompt:** _"Place the product in a modern kitchen setting with natural morning light from the left. Maintain product proportions and add appropriate cast shadows."_ **Why It Works** Context shots require shadow instructions. Products floating without shadows look composited rather than photographed. This prompt creates believable lifestyle imagery by specifying environment, lighting, and shadows. ### **17. Product Reflection Enhancement** **Prompt Structure** - Subject: Product and reflective surface beneath - Action: Add reflection with controlled strength - Preservation: Product color accuracy - Style: Subtle, professional reflection **Example Prompt:** _"Add subtle reflective surface beneath the product with accurate mirror reflection. Keep reflection strength at 15-20 percent and preserve product color accuracy."_ **Why It Works** Percentage controls for reflections prevent over-stylized results that distract from the product itself. A 15-20% reflection adds polish without becoming the focal point. You get professional product presentation that enhances without overwhelming. ## **Restoration and Recovery Prompts** Restoration work brings damaged or aged photos back to usable condition. These prompts handle both physical damage and technical limitations. ### **18. Vintage Photo Restoration** **Prompt Structure** - Subject: Damaged vintage photograph - Action: Repair tears, remove stains, correct fading - Preservation: Original grain character, period authenticity - Style: Restored but historically authentic **Example Prompt:** _"Restore vintage photograph by repairing tears, removing stains, and correcting fading. Preserve original grain character and avoid modern processing artifacts."_ **Why It Works** The grain preservation instruction maintains period authenticity. Over-processing destroys the character that makes vintage photos valuable. You get a clean, usable image that still feels historically accurate. ### **19. Colorize Black and White** **Prompt Structure** - Subject: Black and white photograph - Action: Add realistic color - Preservation: Original shadow detail, period-appropriate tones - Style: Historically accurate colorization **Example Prompt:** _"Add realistic color to black and white photograph using historically accurate tones for the apparent era. Preserve original shadow detail and avoid oversaturation."_ **Why It Works** Historical accuracy matters for restoration clients. Anachronistic colors (like bright modern hues on a 1940s photo) undermine the emotional connection to the image. This prompt adds color while respecting the photograph's era. ### **20. Old Photo Damage Repair** **Prompt Structure** - Subject: Physical damage (scratches, folds, missing sections) - Action: Repair with texture-matched filling - Preservation: All undamaged areas completely untouched - Style: Seamless repair matching original **Example Prompt:** _"Repair physical damage including scratches, fold lines, and missing corners. Match surrounding texture and tone for seamless integration. Preserve undamaged areas completely."_ **Why It Works:** Damage repair prompts need explicit preservation of undamaged areas. Without it, the model might over-process sections that need no work. You get targeted repair that fixes problems without altering what's already good. ### **21. Faded Photo Recovery** **Prompt Structure** - Subject: Faded photograph - Action: Recover contrast and color - Preservation: Natural skin tones, avoid crushed blacks or blown whites - Style: Restored original tonal range **Example Prompt:** _"Recover contrast and color from faded photograph, restoring original tonal range while avoiding crushed shadows or blown highlights. Maintain natural skin tones."_ **Why It Works:** Recovery differs from enhancement. Your goal is returning to original quality, not improving beyond what existed. This prompt brings back lost contrast and color while maintaining realistic tonal balance. ## **Creative and Artistic Prompts** Creative prompts transform images into stylized versions for specific aesthetic goals. These work for social content, artistic projects, and campaign exploration. ### **22. Cinematic Color Grading** **Prompt Structure** - Subject: Overall image color - Action: Apply specific color grade (teal shadows, warm highlights) - Preservation: Natural skin tones, midtone clarity - Style: Cinematic with controlled film grain **Example Prompt:**_ "Apply cinematic color grading with teal shadows and warm highlights at moderate strength. Keep skin tones natural, increase midtone clarity, and add light film grain."_ **Why It Works:** The teal-and-orange color grade remains popular for social content. Specifying skin tone preservation prevents the unnatural color casts that ruin the effect. You get that cinematic look while keeping people looking like people. ### **23. Black and White Conversion** **Prompt Structure** - Subject: Color photograph - Action: Convert to black and white with rich tonal range - Preservation: Highlight and shadow detail - Style: Full tonal range with texture enhancement **Example Prompt:** _"Convert to black and white with rich tonal range from deep blacks to clean whites. Enhance texture detail and add subtle film grain. Preserve highlight and shadow detail."_ **Why It Works:** Quality black and white conversion requires tonal range instructions. Default conversions often lack the depth that makes monochrome images compelling. This prompt creates dramatic, full-range black and white images rather than flat gray conversions. ### **24. Style Transfer** **Prompt Structure** - Subject: Entire image - Action: Apply specific photographic style (film stock, era, technique) - Preservation: Subject detail, natural composition - Style: Specific reference (Kodachrome film characteristics) **Example Prompt:** _"Apply the visual style of vintage Kodachrome film including characteristic color palette, contrast curve, and grain structure. Preserve subject detail and natural composition."_ **Why It Works:** Style transfer prompts benefit from specific reference points. "Vintage" is subjective and could mean anything from 1950s to 1990s. "Kodachrome" has identifiable characteristics (warm color palette, rich saturation, fine grain) that produce consistent results. ### **25. Outfit and Wardrobe Change** **Prompt Structure** - Subject: Specific clothing item - Action: Change color - Preservation: Fabric texture, wrinkles, lighting, face, all other elements - Style: Natural fabric appearance **Example Prompt:**_ "Change the subject's shirt color from blue to deep burgundy while preserving fabric texture, wrinkles, and lighting. Keep face and all other elements identical."_ **Why It Works:** Wardrobe changes require explicit identity preservation. Without it, faces can drift during the edit, creating unusable results. This prompt changes only what you want (shirt color) while locking everything else in place. ## **Integrating AI Editing Into Professional Workflows** These prompts become more valuable when they fit into broader creative processes. A single edit rarely completes a project. Campaigns require variations, approvals, revisions, and format adaptations. When retouching work needs to scale across dozens of images with consistent treatment, the prompt becomes a template. Document what works and apply it systematically. For teams using [Luma's Skills feature](https://lumalabs.ai/news/creative-teams-use-ai), these repeatable workflows can be saved and triggered whenever similar projects arrive. The approval process also shapes how you structure prompts. When clients need to see options, generate variations with controlled differences. When creative direction is locked, use tight preservation rules to prevent drift from approved looks. For images that need ongoing refinement across campaign phases, Luma's [Uni-1 multimodal model](https://lumalabs.ai/uni-1) helps maintain consistency by understanding both visual and text instructions. Each revision builds on previous work rather than starting over. ## **Why Luma for Photo Editing Workflows** When you're managing photo editing at scale, you need tools that understand the full context of your work. Luma's platform integrates AI editing into broader creative workflows rather than treating each edit as an isolated task. The [Layers feature](https://lumalabs.ai/news/introducing-layers) lets you work non-destructively, changing one element while preserving everything else. This matches how professional retouchers actually work, where you need to adjust lighting without touching skin texture, or change backgrounds without affecting your subject. For teams with repeatable workflows, [Skills](https://lumalabs.ai/news/creative-teams-use-ai) let you save proven prompts and apply them consistently across projects. When you develop a portrait retouching approach that works, you can codify it and use it every time similar work arrives. The [Uni-1 model](https://lumalabs.ai/uni-1) brings multimodal understanding to your edits, processing both your visual inputs and text instructions together. You get more consistent results because the model understands context, not just isolated commands. [Try These Prompts Now](https://auth.lumalabs.ai/sign-up) ## **Frequently Asked Questions** ### **What are Gemini AI photo editing prompts?** Gemini AI photo editing prompts are natural language instructions that direct Google's Gemini model to make specific changes to uploaded images. Effective prompts include four key elements: subject identification, action specification, preservation rules, and style parameters. The specificity of your prompt directly affects the quality and predictability of results. ### **Can Gemini AI be used for professional photo retouching?** Yes, with appropriate prompt structure. Professional use requires explicit quality parameters, preservation rules, and specific vocabulary. The prompts in this guide follow professional retouching standards including texture preservation, natural skin tone maintenance, and controlled enhancement levels. Results work best when combined with professional review and refinement workflows. ### **How do I prevent face changes when editing portraits?** Include explicit identity preservation phrases in every portrait prompt. Phrases like "keep face identical," "preserve facial structure," and "maintain exact facial features" instruct the model to treat the face as unchangeable. Without these instructions, edits can cause subtle changes to facial features that make images unusable for your projects. ### **What kind of images benefit most from AI retouching?** High-volume projects with consistent treatment needs see the greatest efficiency gains. Product catalogs, headshot series, event photography, and social media variants all benefit from structured AI editing prompts. Single hero images that require extensive creative judgment still benefit from human direction, with AI handling repetitive technical tasks. ### **How can creative teams integrate Gemini prompts into their workflow?** Start by documenting prompts that produce approved results. Create prompt templates for recurring project types. Build review checkpoints where human editors verify AI output before client delivery. For teams managing multiple campaigns, tools that preserve context across revisions prevent the restart problem that wastes production time.