50+ Gemini AI Photo Prompts
Copy & Paste — Every Use Case
Organised into 10 categories. Click any Copy button to copy the prompt instantly to your clipboard.
Open gemini.google.com or the Gemini app. For editing prompts, upload your photo first (click the image icon), then paste the prompt. For generation prompts, paste directly — no upload needed. Replace any [bracketed text] with your specific details. Add “keep my face exactly as it is” to any portrait prompt to preserve your features.
The Prompt Formula (Why These Work)
Every strong Gemini photo prompt combines these six elements. You don’t need all six every time — but knowing each one lets you fix any prompt that isn’t producing the result you want.
① Subject
Who or what is in the image — a person, product, place.
② Environment
Where the scene takes place — rooftop, studio, forest, city street.
③ Lighting
Quality and direction — golden hour backlight, neon reflections, studio strobe.
④ Camera Style
How it was shot — 85mm f/1.4, wide-angle, shallow depth of field, drone.
⑤ Mood
The feeling — melancholic, warm, cinematic tension, joyful.
⑥ Visual Style
The aesthetic — film noir, Kodak Portra grain, editorial, hyperrealistic 8K.
Portrait Prompts
6 prompts · upload a selfie firstCouple Photo Prompts
4 prompts · upload both peopleFamily Photo Prompts
4 promptsCinematic & Creative Prompts
6 promptsPhoto Editing & Enhancement Prompts
7 prompts · upload your photo firstNature & Landscape Generation Prompts
4 prompts · no upload neededProduct Photography Prompts
4 prompts · upload product photoFantasy & Artistic Prompts
5 promptsArchitecture & Travel Prompts
4 promptsSocial Media & Trending Style Prompts
6 promptsAdvanced Prompting Tips
🔁 Character Consistency
Establish your subject clearly in the first prompt (hair colour, features, clothing). Then use “the same character as before” in follow-up prompts within the same Gemini session to maintain consistency across multiple images.
🧱 Build in Layers
Don’t put everything in one prompt. Start with the most important change, approve it, then add the next layer. Background → lighting → colour grade → grain. Each step is easier to fix than a complex prompt gone wrong in multiple ways.
😶 Face Preservation
If Gemini changes your face, add: “keep my face exactly as it is in the original” and “do not alter facial structure, facial features, or expression.” If it still changes, send a follow-up: “Restore the subject’s face from the reference photo.”
📷 Photography Terms
Use real photo terms: bokeh (blurred background), f/1.4 (shallow depth of field), 85mm (flattering portrait lens), golden hour (warm sunset light), rim light (glow behind subject), Kodak Portra (warm film look), colour grade (overall palette).
🖼️ Multi-Image Prompts
Upload two reference images and combine elements: “Place the person from the first image into the environment from the second image, with natural lighting that matches the environment.” Great for composite portraits and travel photos.
✅ Specificity Wins
“Dramatic lighting” → mediocre. “Single candle light from below, deep shadows under the cheekbones, warm orange tones, shallow depth of field” → cinematic. Pick one element and describe it precisely. Specificity always beats length.
Frequently Asked Questions
Do I need to upload a photo for these prompts?
Editing prompts (categories 5, and portrait/couple/family prompts) require an uploaded photo. Generation prompts (nature, architecture, many cinematic prompts) do not — paste them directly and Gemini creates the image from scratch.
Why does Gemini keep changing my face?
Add “keep my face exactly as it is in the original” or “do not alter facial structure or features” to any portrait prompt. If it still happens, send a follow-up in the same chat: “Restore the face from the reference photo, keep only the [specific change].”
Are these prompts free to use?
Yes, completely free. Gemini has a free tier at gemini.google.com. Gemini Advanced (Google One AI Premium) gives higher quality and more daily generations.
What is the best Gemini prompt structure?
Subject + Environment + Lighting + Camera Style + Mood + Visual Style. You don’t need all six every time. Specificity about one element outperforms vagueness across all six.
Can I use these prompts on other AI tools?
Yes. These prompts are written in a universal photography language that works on ChatGPT (DALL-E 3), Midjourney, Adobe Firefly, and other AI image tools. Some terminology may need minor adjustment for specific platforms.