The prompt library is a curated collection of prompts that have been tested against the models people actually use day to day — ChatGPT, Claude, Gemini and Midjourney among them. Each entry is written to be pasted in as-is, with the placeholders clearly marked so you can swap in your own topic, tone or audience.
Prompts are grouped by the job they do rather than by model: writing and editing, research and summarising, coding and debugging, image generation, and marketing. Most transfer cleanly between models, because the structure that makes a prompt work — clear role, explicit constraints, an example of the output you want — is the same everywhere.
Filter by category, or search for the task you have in mind. Every prompt shows the kind of output it produces before you commit to it.
Copy the prompt and replace the bracketed placeholders with your specifics. The more concrete your details, the less editing the output needs.
Paste it into your model of choice, or run it straight away in TomatoAI Chat. If the result is close but not right, adjust one constraint at a time rather than rewriting the whole prompt.
Yes. Every prompt in the library is free to copy, edit and use commercially. No account is needed to browse or copy them.
They are written to be model-agnostic and work with ChatGPT, Claude, Gemini, Llama and most other chat models. Image prompts are tuned for Midjourney and Stable Diffusion but usually transfer to other image generators with minor edits.
Language models are probabilistic, so the same prompt produces different wording each time. If you need consistency, add explicit constraints — a word count, a format, a worked example of the output you want.
Give the model a role, state the task plainly, list your constraints, and show one example of a good answer. Most weak prompts fail because they leave the format and the audience implicit.
A curated directory of 0 high-quality image generation prompts for Gemini 3 and Nano models.