---
id: 20260825-T0-01
title: "Nexus：同一内存上为Agent模型拼接KV缓存，工具调用提速"
title_en: "Nexus Speeds Up Agentic LLMs with Depth-Adaptive KV-Cache and Tool Routing"
url: https://ai.daily.yangsir.net/daily/20260825-T0-01
issue_date: 2026-08-25
publish_date: 2026-08-24T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2608.20397
---

# Nexus：同一内存上为Agent模型拼接KV缓存，工具调用提速

Nexus是一种针对MCP协议下Agent大语言模型的优化方法，通过在统一内存上拼接深度自适应KV缓存，并解耦检索与工具路由，大幅缩短了工具调用时的首字延迟。在MCP框架中，模型需要不断重新编码冗长的工具定义，导致预填充计算量随序列长度平方增长，成为主要瓶颈。Nexus直接优化了该过程，适用于工具规模庞大、频繁调用外部API的Agent应用，能显著降低响应等待时间。

## English Version

**Nexus Speeds Up Agentic LLMs with Depth-Adaptive KV-Cache and Tool Routing**

Nexus is a novel method for optimizing agentic LLMs in MCP environments, addressing the quadratic prefill bottleneck that prolongs time-to-first-token as tool registries grow. It employs depth-adaptive KV-cache splicing and decouples retrieval from tool routing on unified memory. This reduces overhead from re-encoding verbose tool schemas, enabling faster response times for agents that frequently call external APIs. Nexus is particularly effective for large-scale tool ecosystems, promising significant latency reductions in complex agent workflows.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2608.20397)

**详情页**：https://ai.daily.yangsir.net/daily/20260825-T0-01

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*