---
id: 20260320-T0-12
title: "无需嵌入，构建知识代理的新方案"
title_en: "Build Knowledge Agents Without Embeddings"
url: https://ai.daily.yangsir.net/daily/20260320-T0-12
issue_date: 2026-03-20
publish_date: 2026-03-19T16:00:00.000Z
category: insight
source_name: "Vercel Blog"
source_url: https://vercel.com/blog/build-knowledge-agents-without-embeddings
---

# 无需嵌入，构建知识代理的新方案

Vercel提出知识代理的新构建方式。传统方法需选择向量数据库、分词管道和嵌入模型，耗时且难以调试。新方案绕过嵌入步骤，直接从知识库检索信息，大幅简化流程。开发者无需手动调整检索参数，代理能更准确地回答问题，并可追溯检索来源。该方法适用于需要高精度知识检索的场景。

## English Version

**Build Knowledge Agents Without Embeddings**

Vercel introduces a new approach to building knowledge agents that skips embeddings. Traditional methods require vector databases, chunking pipelines, and embedding models, which are time-consuming and hard to debug. The new approach retrieves information directly from knowledge bases, eliminating embedding steps. This simplifies development, improves answer accuracy, and allows for retrieval traceability, making it ideal for high-precision knowledge retrieval scenarios.

---

**来源**：[Vercel Blog](https://vercel.com/blog/build-knowledge-agents-without-embeddings)

**详情页**：https://ai.daily.yangsir.net/daily/20260320-T0-12

---

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