AI Blocks › Make product photos
Product background swapper
Give it a white-background product photo. Get it placed in a real-looking scene with matching light. Your image model draws it.
Product background swapper is a free AI skill for the job "Make product photos". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · MIT licence · Made by dLazy · Level: beginner · Thinking: low
Made by dLazy. See the original on GitHub; we copied it as it is, with its licence.
What it does
Your plain white-background product photo placed into a believable room, desk or outdoor scene. The product stays exactly as it is. Only the surroundings change. The method focuses on three things that make a swap look real: a contact shadow under the product, reflections from nearby surfaces, and light that comes from the same direction as in the scene. Without these the product looks like a sticker floating on a backdrop. Claude writes the scene prompt that keeps the product details fixed and describes the setting and light. Claude cannot draw the picture itself. An image model such as ChatGPT or Gemini, or an image API key, does that step.
You give it
- A white-background product photo
- The scene you want, such as a wooden desk by a window
- Mood or time of day (optional)
- Access to an image model
You get
- A ready-to-use image prompt
- A scene with grounded shadows and matching light
- Notes on keeping the product unchanged
- Tips to fix a swap that looks fake
Use it when
- You have clean white-background shots and want lifestyle images for ads or listings.
- A product photo looks flat and you want it on a table, floor or shelf.
- You need the same product shown in several seasons or settings.
- A stock backdrop made your product look pasted on.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can run code and use your own image model key (OpenAI, Gemini, fal or Replicate).