How it works
Last updated October 2026
ShootKits prompts turn one photo of your product into a professional shoot in ChatGPT. You upload the product, paste the prompt, and ChatGPT creates the set, light, model and styling while keeping your product exactly as it is.
The three steps
- Pick a kit. Browse by sector or by style. Each kit shows the finished result.
- Upload your product. In ChatGPT, attach a clear photo of your product before you paste anything.
- Paste the prompt and send. Campaign kits return a grid of shots; single shots return one photo.
How do I make product photos with ChatGPT?
Open ChatGPT, attach a clear photo of your product, paste a ShootKits prompt and send it. ChatGPT keeps your product as it is and builds the set, lighting and model around it.
What kind of product photo should I upload?
Upload a sharp photo of the product alone on a plain background, in soft even light, showing its front and label. A good phone photo works; avoid strong shadows, reflections and clutter.
Why does ChatGPT change my product, and how do I fix it?
Image models sometimes simplify small details like logos, text or stitching. Reply in the same chat with exactly what to keep, for example “Keep the label text exactly as in my photo”, and ChatGPT regenerates with that fix.
What is the difference between a campaign kit and a single shot?
A campaign kit makes a grid of six or nine related shots in one go, with the same product and model across every panel. A single shot makes one finished photo.
Can I make one panel of a campaign at full size?
Yes. Every campaign kit has a shot list, and each shot has its own prompt. Copy the shot you want and send it with your product photo to get that panel as a full-size image.
Can I use the images commercially?
The prompts are free to use. Rights to images you generate are set by OpenAI’s terms for ChatGPT, so check their current usage policy for your plan before using images in ads or on product listings.
Do the prompts work with other AI image tools?
They are written and tested for ChatGPT. Most also work in other image models that accept a reference photo, but results vary and we only test in ChatGPT.
Tips for better results
- One product per photo. Extra items confuse the model about what to keep.
- Show the side that matters. If the label or dial is the point, photograph it straight on.
- Keep the chat going. Fixes in the same conversation keep everything else the same.
- Use the shot list. Generate your favourite panels one by one for higher resolution.
- Check text and logos. Always zoom in on labels before you publish an image.