Unverified50% confidenceFactExact time
An Agent's token consumption grows linearly or even superlinearly with the number of conversation turns and tool calls.
1
Sources
50%
Confidence
Long-term
Relevance
10/24/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAgent框架慢于直接对话主要因为注入大量系统提示词工具定义、ReAct多轮推理和上下文累积导致输入token数增长79% similarUnverifiedAgentExecutor每次工具调用前会重新拼接完整对话历史进Prompt，导致Token消耗随任务步骤线性增长78% similarUnverifiedWhen using an Agent, multi-turn dialogue should be treated as part of the workflow, not expecting a perfect delivery in one go; the interaction process itself consumes credits, which is a normal cost of using an Agent.78% similarUnverifiedIn agent scenarios, the context window accumulates continuously with each conversation turn, causing the input tokens for each subsequent turn to grow persistently.77% similarUnverifiedIn agentic scenarios, an Agent often initiates multiple rounds of model calls within a single user request, with each round carrying an increasingly longer context, causing costs to grow nonlinearly with conversation depth.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1092170API
curl https://kongchang.com/api/v1/knowledge/claims/1092170MCP
get_claim(id=1092170)