Unverified50% confidenceFactExact time
In ReAct-style Agents, each loop iteration must reinsert the historical thoughts and tool results into the context, so the more rounds, the higher the token consumption and cost.
1
Sources
50%
Confidence
Long-term
Relevance
10/24/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn each ReAct cycle, the complete history context must be re-input, causing the context window to grow linearly with the number of iterations.80% similarUnverifiedReAct's reasoning chain continuously consumes context tokens; for an agent executing 20 steps, the accumulated reasoning history may consume tens of thousands of tokens.79% similarUnverified典型的Agent框架如ReAct、AutoGPT、LangGraph采用感知-推理-行动的迭代模式，上下文随轮次累积导致后期每轮Token消耗呈线性甚至指数增长75% similarUnverifiedIn classic paradigms such as ReAct, tool-call decisions are made by the same large model, and when the task chain is long the loop can run dozens to hundreds of rounds, with latency and cost accumulating linearly.74% similarUnverifiedThe cost of ReAct is that each loop adds one model call, with latency and token consumption growing linearly with the number of steps, so real systems typically set a maximum number of iterations and early-termination conditions.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1092168API
curl https://kongchang.com/api/v1/knowledge/claims/1092168MCP
get_claim(id=1092168)