Choose a model, write a prompt and adjust generation settings for your task.
Compare output type, reference support, available settings and price before generating.
A simple request structure, examples of improvements and prompts you can copy.