system prompt
/ˈsɪstəm ˌprɒmpt/
high-priority instructions supplied by an application or platform to shape model behavior; the role name, visibility, and scope vary by system
system prompt in a sentence
“The application used a system-level instruction to request concise, formal responses.”
Origin of system prompt
Greek systēma organized whole + Latin promptus
What does system prompt really mean?
A system prompt is a common name for high-priority instructions supplied by an application or platform. It can define a role, constraints, available tools, and response style, but it is not a universal or permanent layer: APIs differ in role names, priority rules, visibility, and how long instructions remain in scope.
The story behind system prompt
The term became common with role-tagged chat APIs, where messages could be labeled system, user, or assistant. The idea is older than the name: completion systems already used instruction and persona preambles. Some current APIs use a developer role or a separate system parameter instead, so the exact label is provider-specific.
How to use system prompt
Use it when a product genuinely has a system-level instruction layer: "the response style came from the system prompt." Do not assume that layer is hidden, conversation-long, or a security boundary. Prompt injection can arrive directly from a user or indirectly through files, webpages, and tool results; applications still need permissions, isolation, and validation.
Related Words
few-shot learning
providing examples in the prompt to demonstrate desired input-output patterns
zero-shot
asking a model to perform a task without any examples
chain-of-thought
intermediate reasoning steps generated while solving a task, whether elicited by prompting or produced internally by a reasoning model
prompt injection
an attack or failure mode in which instructions embedded in user input or external content steer an LLM application away from its intended behavior
context priming
using early context to set expectations and influence subsequent model behavior
meta-prompting
asking the model to help design or improve prompts for itself