> ## Documentation Index
> Fetch the complete documentation index at: https://cndoc-langchain.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone 集成

> 使用 LangChain Python 与 Pinecone 集成。

> [Pinecone](https://docs.pinecone.io/docs/overview) 是一个功能广泛的向量数据库。

## 安装与设置

安装 Python SDK：

<CodeGroup>
  ```bash pip theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
  pip install langchain-pinecone
  ```

  ```bash uv theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
  uv add langchain-pinecone
  ```
</CodeGroup>

## 向量存储

存在一个围绕 Pinecone 索引的封装器，允许你将其用作向量存储，无论是用于语义搜索还是示例选择。

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
from langchain_pinecone import PineconeVectorStore
```

有关 Pinecone 向量存储的更详细演练，请参阅[此笔记本](/oss/python/integrations/vectorstores/pinecone)

### 稀疏向量存储

LangChain 的 `PineconeSparseVectorStore` 支持使用 Pinecone 的稀疏英文模型进行稀疏检索。它将文本映射到稀疏向量，并支持添加文档和相似性搜索。

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
from langchain_pinecone import PineconeSparseVectorStore

# 初始化稀疏向量存储
vector_store = PineconeSparseVectorStore(
    index=my_index,
    embedding_model="pinecone-sparse-english-v0"
)
# 添加文档
vector_store.add_documents(documents)
# 查询
results = vector_store.similarity_search("your query", k=3)
```

有关更详细的演练，请参阅 [Pinecone 稀疏向量存储笔记本](/oss/python/integrations/vectorstores/pinecone_sparse)。

### 稀疏嵌入

LangChain 的 `PineconeSparseEmbeddings` 使用 Pinecone 的 `pinecone-sparse-english-v0` 模型提供稀疏嵌入生成。

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
from langchain_pinecone.embeddings import PineconeSparseEmbeddings

# 初始化稀疏嵌入
sparse_embeddings = PineconeSparseEmbeddings(
    model="pinecone-sparse-english-v0"
)
# 嵌入单个查询（返回 SparseValues）
query_embedding = sparse_embeddings.embed_query("sample text")

# 嵌入多个文档（返回 SparseValues 列表）
docs = ["Document 1 content", "Document 2 content"]
doc_embeddings = sparse_embeddings.embed_documents(docs)
```

有关更详细的用法，请参阅 [Pinecone 稀疏嵌入笔记本](/oss/python/integrations/vectorstores/pinecone_sparse)。

## 检索器

### Pinecone 混合搜索

<CodeGroup>
  ```bash pip theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
  pip install pinecone pinecone-text
  ```

  ```bash uv theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
  uv add pinecone pinecone-text
  ```
</CodeGroup>

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
from langchain_community.retrievers import (
    PineconeHybridSearchRetriever,
)
```

有关更详细的信息，请参阅[此笔记本](/oss/python/integrations/retrievers/pinecone_hybrid_search)。

### 自查询检索器

Pinecone 向量存储可用作自查询的检索器。

***

<div className="source-links">
  <Callout icon="terminal-2">
    [将这些文档](/use-these-docs)通过 MCP 连接到 Claude、VSCode 等，以获取实时答案。
  </Callout>

  <Callout icon="edit">
    [在 GitHub 上编辑此页面](https://github.com/langchain-ai/docs/edit/main/src/oss/python/integrations/providers/pinecone.mdx) 或 [提交问题](https://github.com/langchain-ai/docs/issues/new/choose)。
  </Callout>
</div>