Unverified50% confidenceFactExact time
在智能体工作流中，历史交互作为输入Token累积，导致总Token消耗随任务轮次呈近似平方级增长
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
GPT-5.6 Three-Model Showdown: Sol vs Terra vs Luna — Which One Should You Pick?
bilibili麦冬AI实验室7/11/2026
Related Claims
UnverifiedWhen an agent is in a loop, each call carries the accumulated conversation history, causing token consumption to grow linearly or even super-linearly with the number of iterations.78% similarUnverified智能体模式下Token消耗呈非线性增长，因每次工具调用结果追加到上下文中，多步骤任务的Token消耗可达O(n²)级别，使低成本模型对智能体场景至关重要76% similarUnverified智能体系统常通过定时任务每分钟不间断运行，加上多轮模型调用、工具调用结果重新输入及反思步骤，令令牌成本呈指数级攀升74% similarUnverified多Agent架构中Agent间传递上下文时重复信息被反复处理，导致Token消耗和成本上升74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/602594API
curl https://kongchang.com/api/v1/knowledge/claims/602594MCP
get_claim(id=602594)