A good prompt is structured: a role, an objective, a context, and constraints. This generator assembles these elements into a ready-to-copy prompt, without sending anything to an AI.
How to use it
- Start from a predefined template or from scratch, and choose the style: structured Markdown, XML tags, or compact text.
- Fill in the role, target audience, tone, output format, length, then the task, context, and constraints.
- Check the request for step-by-step reasoning if needed, then copy the prompt.
The principles behind the calculation
- The prompt is assembled directly in your browser: no text is transmitted to an AI service.
- Only completed sections are kept: role, objective, context, audience, tone, output format, length, and constraints.
- The response language and the step-by-step reasoning instruction are added to the constraints. Brackets, like , mark what remains to be filled in.
- The prompt’s token count is estimated in real time using the same calculation as the token counter.
Limitations to be aware of
The generator structures your request, it does not make it better on its own: the quality depends on the information you provide and the model used.
