Unverified50% confidenceFactExact time
Overly long system prompts bring hidden overhead: for example, when the system prompt has 500 tokens while the actual input has only 50 tokens, it means 91% of the cost goes to boilerplate text; auditing and trimming the system prompt by 40% can reduce costs.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
Unverified对于使用长系统提示、RAG文档注入或代码仓库上下文的企业级应用，显式断点可将Token成本降低40%-70%77% similarUnverifiedIn agentic scenarios where system prompts run to thousands of tokens, if more than 90% of requests hit the cache, the effective input cost per request may be only 5%–10% of the listed price.73% similarVerifiedToken billing is calculated based on the total volume of input and output; the longer a conversation, the more each call pays for an increasingly bloated history.73% similarVerifiedThe /compact command can reduce token consumption by 30% to 50%.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1038720API
curl https://kongchang.com/api/v1/knowledge/claims/1038720MCP
get_claim(id=1038720)