Start from the job you need done, then pick the model. Each page collects prompts that share the same commercial goal.
Pick the model you are generating with, open its hub, and browse the prompt grid. Every prompt page shows the full prompt text with a one-click copy button, the reference image, and the tags it belongs to. Prompts are written for the specific habits of each model, so the wording that works in Midjourney is not the wording that works in GPT Image.
If you are not sure which model to use, start from the use case pages instead: they group prompts by the job you are trying to get done, not by the tool.