Find answers to frequently asked questions about PromptMagic. Browse by category or search for specific topics.
Showing 11-20 of 28 questions
System = rules, User = request, Assistant = AI reply.
They differ by how many examples you give — none, one, or a few.
Lower settings = safer, less creative answers; higher = more variety, possible risk of errors.
Prompt engineering is the practice of carefully crafting and refining inputs, or "prompts," to guide generative AI models toward producing accurate, relevant, and desired outputs.
Split tasks into modules, use bullet lists or sub-questions, and add structure to requests.
No; custom prompts remain critical for precision, safety, and unique use cases.
Use imperative tone, step-by-step logic, and specify penalties for deviation.
Request source-backed answers and require references for every claim.
Specifying output type (table, bullets, code) increases clarity and usability.
Ask for step-by-step breakdowns or “show your work” in the prompt.
If you couldn't find an answer to your question, our support team is here to help.