Unverified50% confidenceFactExact time
A large model is essentially one question and one answer; a single API request itself carries no prior context, and the server does not save conversation context for any particular user.
1
Sources
50%
Confidence
Long-term
Relevance
10/21/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDue to the stateless design of large model APIs, the entire tools array must be resent in every request, alongside the system prompt and conversation history.72% similarUnverifiedThe minimal pipeline of a large model Q&A system is: the user sends parameters from the frontend to the WebApp, which forwards them to the backend API via HTTP request, and the backend calls the large model for processing and renders the result back to the frontend.71% similarUnverifiedAfter receiving a request, the model may return a structured tool call rather than directly providing an answer.70% similarUnverified商业化大模型服务会在用户可见对话之外附加系统提示，用于设定模型角色、行为规范、安全边界及输出风格，系统提示随每次API请求完整发送至模型70% similarUnverifiedThe messages parameter in an API request contains the conversation content, typically including two roles: system and user.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1081899API
curl https://kongchang.com/api/v1/knowledge/claims/1081899MCP
get_claim(id=1081899)