- We write or adapt the prompt and keep the “same face” line at the start.
- We make a sample photo with an AI image tool to show the look the prompt aims for, and label it “Sample result”.
- When we run the prompt on a listed tool ourselves (for example Gemini or ChatGPT), we add that photo with the tool name as its label and show the “Tested” date.
- We update prompts when the tools change or when readers tell us a prompt has stopped working.
How we make prompts
What goes into each prompt and its sample photo.
Last updated 19 Oct 2026