How to Turn Any Photo into an AI Prompt Using Google Gemini & Claude

Published on: August 29, 2026 | By PromptAqua Editorial Team

Have you ever come across a stunning image or portrait online and wondered how to recreate a similar photo using AI? Thanks to multi-modal AI assistants like Google Gemini and Anthropic Claude, reverse-engineering image prompts has become effortless. In this comprehensive guide, we show you how to extract detailed prompts from photos and generate custom visual styles.

Step 1: Uploading Your Reference Photo to Gemini or Claude

Both Google Gemini and Claude allow image uploads. Follow these initial steps:

Step 2: Using the Master Reverse-Prompting Instruction

Once the photo is attached, copy and paste this exact prompt template into the AI chat box:

"Analyze this uploaded image in detail. Describe the main subject, lighting setup, color palette, camera angle, depth of field, and art style. Then, write a detailed 100-word text prompt optimized for Midjourney v6 and Flux.1 to recreate a similar aesthetic."

Step 3: Generating Your Custom AI Image

Gemini or Claude will provide a detailed prompt breakdown. You can now:

Pro Tip for Converting Video Clips into Prompts

If you have a video clip, simply pause at the best frame, take a screenshot, and upload that screenshot to Gemini. You can extract cinematic lighting prompts from videos within seconds!

Conclusion

Combining Vision AI models with image generators gives you complete creative freedom. Explore pre-tested, high-quality prompts on our PromptAqua Home Page to supercharge your design workflow!