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

# 嵌入模型集成

> 集成嵌入模型，使用LangChain Python。

## 概述

> [!NOTE]
>
> 本概述涵盖 **基于文本的嵌入模型**。LangChain 目前不支持多模态嵌入。
>
> 参见 [顶级嵌入模型](#顶级集成)。

嵌入模型将原始文本（如句子、段落或推文）转换为固定长度的数字向量，该向量捕获其 **语义意义**。这些向量使机器能够根据意义而不是确切词语来比较和搜索文本。

在实践中，这意味着具有相似想法的文本在向量空间中会被放置在一起。例如，嵌入模型不仅可以匹配 *"机器学习"* 这个短语，还可以在措辞不同的情况下提取讨论相关概念的文档。

Integration components
Arrows show a typical retrieval flow. The highlighted component is the one this page covers.

### Typical retrieval flow

### Agent capabilities

### Execution and state

### 工作原理

1. **向量化**：模型将每个输入字符串编码为高维向量。
2. **相似度评分**：使用数学指标比较向量，以衡量底层文本的关联程度。

### 相似度指标

通常使用以下指标来比较嵌入：

* **余弦相似度**：测量两个向量之间的夹角。
* **欧几里得距离**：测量两点之间的直线距离。
* **点积**：测量一个向量在另一个向量上的投影程度。

以下是计算两个向量之间余弦相似度的示例：

```python theme={null}
import numpy as np

def cosine_similarity(vec1, vec2):
    dot = np.dot(vec1, vec2)
    return dot / (np.linalg.norm(vec1) * np.linalg.norm(vec2))

similarity = cosine_similarity(query_embedding, document_embedding)
print("Cosine Similarity:", similarity)
```

## 接口

LangChain 通过 @\[Embeddings] 接口为文本嵌入模型（如 OpenAI、Cohere、Hugging Face）提供标准接口。

提供两种主要方法：

* `embed_documents(texts: List[str]) → List[List[float]]`：嵌入文档列表。
* `embed_query(text: str) → List[float]`：嵌入单个查询。

> [!NOTE]
>
> 该接口允许使用不同的策略嵌入查询和文档，尽管在实践中大多数提供商会以相同的方式处理它们。

## 顶级集成

| 集成 | 下载 |
| :- | :- |
| [`AzureOpenAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/azure_openai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-openai/month)](https://pypi.org/project/langchain-openai/) |
| [`OpenAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/openai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-openai/month)](https://pypi.org/project/langchain-openai/) |
| [`GoogleGenerativeAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/google_generative_ai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-google-genai/month)](https://pypi.org/project/langchain-google-genai/) |
| [`OllamaEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/ollama) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-ollama/month)](https://pypi.org/project/langchain-ollama/) |
| [`HuggingFaceEmbeddings (Sentence Transformers)`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/sentence_transformers) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-huggingface/month)](https://pypi.org/project/langchain-huggingface/) |
| [`MistralAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/mistralai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-mistralai/month)](https://pypi.org/project/langchain-mistralai/) |
| [`CohereEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/cohere) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-cohere/month)](https://pypi.org/project/langchain-cohere/) |
| [`NVIDIAEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/nvidia_ai_endpoints) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-nvidia-ai-endpoints/month)](https://pypi.org/project/langchain-nvidia-ai-endpoints/) |
| [`DatabricksEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/databricks) | [  ![Downloads per month](https://static.pepy.tech/badge/databricks-langchain/month)](https://pypi.org/project/databricks-langchain/) |
| [`PerplexityEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/perplexity) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-perplexity/month)](https://pypi.org/project/langchain-perplexity/) |
| [`TogetherEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/together) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-together/month)](https://pypi.org/project/langchain-together/) |

### 常见的部署模式

在实践中，大多数团队会采用以下四种模式之一：

1. 托管旗舰：OpenAI `text-embedding-3-large`、Cohere `embed-english-v3`、Google `gemini-embedding-001`、Voyage `voyage-3`。一个 API 调用，开箱即用的最佳质量，无需本地基础设施。按调用付费，存在数据外传依赖。
2. 本地开源：`BAAI/bge-*`、`mixedbread-ai/mxbai-embed-*`、`Qwen/Qwen3-Embedding-*`、`nomic-ai/modernbert-embed-*`、`sentence-transformers/all-*`。下载一次，随处运行。无按调用付费，数据不会离开您的环境。在 CPU 上规模较小的情况下可能比托管 API 慢；使用 GPU 时具有竞争力或更快。
3. 本地开源，专业型：针对您的特定领域、语言或任务进行微调的模型。从强大的开源基础（例如 `BAAI/bge-m3`）开始，并在数千个领域内查询/文档对上进行微调，通常在检索准确性方面优于该领域的托管旗舰。
4. 生产规模自托管：通过 [文本嵌入推理 (TEI)](https://github.com/huggingface/text-embeddings-inference) 或 Ollama 服务的相同开源模型（基础或微调）。您获得本地推理的经济效益，以及托管提供者的水平扩展和 API 便利性。

LangChain 对这四种模式一视同仁：您实例化一个 `Embeddings` 子类，并将其交给您的向量存储或检索器。模式 (2) 和 (3) 使用 `HuggingFaceEmbeddings`；模式 (4) 使用 `OpenAIEmbeddings` 对 TEI 的 OpenAI 兼容端点，或 `OllamaEmbeddings`。

### 需要权衡的因素

#### 质量

从 [MTEB 排行榜](https://huggingface.co/spaces/mteb/leaderboard) 开始。MTEB 跨检索、聚类、分类和重新排序任务对嵌入模型进行基准测试，是事实上的行业参考。按您的语言和任务（检索是最常见的 RAG 任务）进行筛选。

排行榜数字并不总是能直接转换，因此在承诺之前，请在自己的数据上运行小型评估。LangSmith 提供了用于此目的的工具；参见 [评估指南](https://docs.langchain.com/langsmith/evaluation-concepts)。

#### 成本

托管嵌入通常按百万 Token 几分钱到 \~\$0.15 的价格定价。对于一次嵌入并每天查询数千次的语料库，成本通常由查询端主导。

本地推理没有按调用成本，但需要 CPU（慢）或 GPU（资本或云成本）。交叉点取决于工作负载：低容量的个人项目在 CPU 上几乎免费；对于中等容量的生产环境，通过 TEI 服务的单个 GPU 本地模型通常在单位经济性上优于托管服务。

#### 延迟

托管嵌入 API 每个请求大约增加 50-200ms 的网络延迟。CPU 上的本地模型对于包含小模型的短查询（`all-MiniLM-L6-v2`-类）需要 10-100ms，而对于较大的模型需要 50-500ms。在 GPU 上，本地推理通常比往返托管 API 更快。

对于批量索引，每个请求的延迟不如吞吐量重要。TEI 和多进程本地推理会激进地批量处理。在 GPU 上运行时，考虑 `encode_kwargs={"batch_size": 64}` 或更高的 `HuggingFaceEmbeddings`。

#### 维度

嵌入维度影响向量存储存储和查询计算。典型大小：

* 384（小 Sentence Transformers 模型，`all-MiniLM-L6-v2`）
* 768（中等 Sentence Transformers 模型，`all-mpnet-base-v2`、`bge-base`）
* 1024（`bge-large`、Cohere v3、Voyage）
* 1536（OpenAI `text-embedding-3-small`、Qwen3-Embedding-0.6B）
* 3072+（OpenAI `text-embedding-3-large`、Qwen3-Embedding-4B/8B）

较大的向量通常更准确，但会消耗更多存储和查询计算。一些现代模型（OpenAI `text-embedding-3-*`、`mixedbread-ai/mxbai-embed-large-v1`、Matryoshka 训练的 Sentence Transformers 模型、Qwen3-Embedding）支持 **截断**：将向量切片到更小的维度，同时优雅地降低质量。这对于将更多向量拟合到更小的索引中很有用。

#### 上下文长度

大多数经典嵌入模型在 512 个 Token（`all-mpnet-base-v2`、经典 BGE）上达到上限。更新的模型支持更长的上下文：

* `nomic-ai/modernbert-embed-base`：8192 个 Token
* `Alibaba-NLP/gte-multilingual-base`：8192 个 Token
* `BAAI/bge-m3`：8192 个 Token
* OpenAI `text-embedding-3-*`：8191 个 Token

如果您的片段很长（全页技术文档、法律段落），请优先选择长上下文模型。对于短片段，512 个 Token 的限制很少具有约束性。

#### 多语言支持

对于多语言检索，请选择在您的语言上训练的模型。强大的默认选项：

* 开源：`BAAI/bge-m3`、`intfloat/multilingual-e5-*`、`Alibaba-NLP/gte-multilingual-*`、`Qwen/Qwen3-Embedding-*`（通过 `HuggingFaceEmbeddings`）
* 托管：Cohere `embed-multilingual-v3`、OpenAI `text-embedding-3-*`

#### 查询和文档提示

几个现代开源模型（E5、BGE、Qwen3-嵌入、GTE）在查询和文档上使用不同的文本前缀进行训练。在查询时使用错误的前缀是一个常见的质量退化问题。当使用 `HuggingFaceEmbeddings` 时，请显式传递提示词：

```python theme={null}
from langchain_huggingface import HuggingFaceEmbeddings

embeddings = HuggingFaceEmbeddings(
    model_name="intfloat/e5-large-v2",
    encode_kwargs={"prompt": "passage: "},
    query_encode_kwargs={"prompt": "query: "},
)
```

检查 Hugging Face 上的每个模型的卡片以获取推荐的提示词字符串。

#### 许可证

大多数流行的开源嵌入模型采用宽松的许可证（Apache 2.0、MIT）。一些最近的专用模型在生产使用时需要商业许可证。在发布前请检查每个模型的许可证。

### 超越单个向量密集嵌入

每个片段一个密集向量是默认选项，但不是唯一选项。

#### 稀疏和混合检索

密集嵌入不如基于关键词的索引处理精确匹配查询（产品代码、命名实体、代码标识符）那么好。混合检索结合了密集索引与 BM25 或稀疏神经索引（SPLADE、`BAAI/bge-m3` 的稀疏输出）来覆盖这两种情况。

#### 晚交互和多向量

ColBERT 风格的模型为每个标记而不是每个片段生成一个向量，然后通过晚交互对查询进行评分。这在处理复杂查询时通常比单个向量密集检索更准确，但代价是更高的存储需求和更复杂的索引。这个领域的当前开源模型包括 `jinaai/jina-colbert-v2`、`answerdotai/answerai-colbert-small-v1` 以及新的晚交互变体如 `lightonai/LateOn`。LangChain 的内置检索器针对单个向量嵌入；晚交互通常需要一个专用索引（Vespa、Qdrant 的多向量支持或 PyLate）。

### 启动点

如果你只需要一个工作的起点：

* 快速原型，托管：`OpenAIEmbeddings(model="text-embedding-3-small")`
* 快速原型，本地，无 API 密钥：`HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2", encode_kwargs={"normalize_embeddings": True})`
* 生产，托管，质量优先：`VoyageAIEmbeddings(model="voyage-3")` 或 `OpenAIEmbeddings(model="text-embedding-3-large")`
* 生产，开源，质量优先：`HuggingFaceEmbeddings(model_name="BAAI/bge-m3", encode_kwargs={"normalize_embeddings": True})` 通过 TEI 提供
* 多语言，开源：`HuggingFaceEmbeddings(model_name="intfloat/multilingual-e5-large")`，配置查询和文档提示词

在你的数据上测量检索质量，然后进行迭代。

## 缓存

嵌入可以存储或临时缓存以避免需要重新计算它们。

使用 `CacheBackedEmbeddings` 可以进行嵌入缓存。这个包装器将嵌入存储在键值存储中，其中文本被哈希，哈希值用作缓存中的键。

初始化 `CacheBackedEmbeddings` 的主要支持方式是 `from_bytes_store`。它接受以下参数：

* **`underlying_embedder`**：用于嵌入的嵌入器。
* **`document_embedding_cache`**：任何 [`ByteStore`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/stores/) 用于缓存文档嵌入。
* **`batch_size`**：（可选，默认为 `None`）在存储更新之间嵌入的文档数量。
* **`namespace`**：（可选，默认为 `""`）用于文档缓存的命名空间。有助于避免冲突（例如，将其设置为嵌入模型名称）。
* **`query_embedding_cache`**：（可选，默认为 `None`）用于缓存查询嵌入的 [`ByteStore`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/stores/)，或 `True` 以重用与 `document_embedding_cache` 相同的存储。

- 使用不同的嵌入模型时，始终设置 `namespace` 参数以避免冲突。
- `CacheBackedEmbeddings` 默认不缓存查询嵌入。要启用此功能，请指定一个 `query_embedding_cache`。

```python theme={null}
import time
from langchain_classic.embeddings import CacheBackedEmbeddings  # [!code highlight]
from langchain_classic.storage import LocalFileStore # [!code highlight]
from langchain_core.vectorstores import InMemoryVectorStore

# Create your underlying embeddings model
underlying_embeddings = ... # e.g., OpenAIEmbeddings(), HuggingFaceEmbeddings(), etc.

# Store persists embeddings to the local filesystem
# This isn't for production use, but is useful for local
store = LocalFileStore("./cache/") # [!code highlight]

cached_embedder = CacheBackedEmbeddings.from_bytes_store(
    underlying_embeddings,
    store,
    namespace=underlying_embeddings.model
)

# Example: caching a query embedding
tic = time.time()
print(cached_embedder.embed_query("Hello, world!"))
print(f"First call took: {time.time() - tic:.2f} seconds")

# Subsequent calls use the cache
tic = time.time()
print(cached_embedder.embed_query("Hello, world!"))
print(f"Second call took: {time.time() - tic:.2f} seconds")
```

在生产环境中，你通常会使用更健壮的持久存储，例如数据库或云存储。请参阅 [stores integrations](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/stores.md) 获取选项。

## 所有嵌入模型

| 集成 | 下载 |
| :- | :- |
| [`AzureOpenAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/azure_openai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-openai/month)](https://pypi.org/project/langchain-openai/) |
| [`OpenAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/openai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-openai/month)](https://pypi.org/project/langchain-openai/) |
| [`Gemini Enterprise Agent Platform`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/google_vertex_ai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-google-vertexai/month)](https://pypi.org/project/langchain-google-vertexai/) |
| [`GoogleGenerativeAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/google_generative_ai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-google-genai/month)](https://pypi.org/project/langchain-google-genai/) |
| [`BedrockEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/bedrock) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-aws/month)](https://pypi.org/project/langchain-aws/) |
| [`OllamaEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/ollama) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-ollama/month)](https://pypi.org/project/langchain-ollama/) |
| [`BGE on Hugging Face`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/bge_huggingface) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-huggingface/month)](https://pypi.org/project/langchain-huggingface/) |
| [`Hugging Face`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/huggingfacehub) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-huggingface/month)](https://pypi.org/project/langchain-huggingface/) |
| [`HuggingFaceEmbeddings (Sentence Transformers)`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/sentence_transformers) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-huggingface/month)](https://pypi.org/project/langchain-huggingface/) |
| [`Instructor embeddings on Hugging Face`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/instruct_embeddings) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-huggingface/month)](https://pypi.org/project/langchain-huggingface/) |
| [`MistralAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/mistralai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-mistralai/month)](https://pypi.org/project/langchain-mistralai/) |
| [`Text embeddings inference`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/text_embeddings_inference) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-huggingface/month)](https://pypi.org/project/langchain-huggingface/) |
| [`FireworksEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/fireworks) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-fireworks/month)](https://pypi.org/project/langchain-fireworks/) |
| [`CohereEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/cohere) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-cohere/month)](https://pypi.org/project/langchain-cohere/) |
| [`NVIDIAEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/nvidia_ai_endpoints) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-nvidia-ai-endpoints/month)](https://pypi.org/project/langchain-nvidia-ai-endpoints/) |
| [`DatabricksEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/databricks) | [  ![Downloads per month](https://static.pepy.tech/badge/databricks-langchain/month)](https://pypi.org/project/databricks-langchain/) |
| [`Pinecone`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/pinecone) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-pinecone/month)](https://pypi.org/project/langchain-pinecone/) |
| [`WatsonxEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/ibm_watsonx) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-ibm/month)](https://pypi.org/project/langchain-ibm/) |
| [`PerplexityEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/perplexity) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-perplexity/month)](https://pypi.org/project/langchain-perplexity/) |
| [`Oracle AI vector search generate`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/oracleai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-oracledb/month)](https://pypi.org/project/langchain-oracledb/) |
| [`Elasticsearch`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/elasticsearch) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-elasticsearch/month)](https://pypi.org/project/langchain-elasticsearch/) |
| [`Voyage AI`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/voyageai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-voyageai/month)](https://pypi.org/project/langchain-voyageai/) |
| [`BasetenEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/baseten) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-baseten/month)](https://pypi.org/project/langchain-baseten/) |
| [`TogetherEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/together) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-together/month)](https://pypi.org/project/langchain-together/) |
| [`OCIGenAIEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/oci_generative_ai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-oci/month)](https://pypi.org/project/langchain-oci/) |
| [`UpstageEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/upstage) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-upstage/month)](https://pypi.org/project/langchain-upstage/) |
| [`SambanovaEmbeddings`](https://langchain-docs-zh.mintlify.site/langchain/oss/python/integrations/embeddings/sambanova) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-sambanova/month)](https://pypi.org/project/langchain-sambanova/) |
| [`Cloudflare workers AI`](https://developers.cloudflare.com/ai/models/#text-embeddings) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-cloudflare/month)](https://pypi.org/project/langchain-cloudflare/) |
| [`Naver`](https://guide.ncloud-docs.com/docs/clovastudio-dev-langchain) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-naver/month)](https://pypi.org/project/langchain-naver/) |
| [`NomicEmbeddings`](https://atlas.nomic.ai/) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-nomic/month)](https://pypi.org/project/langchain-nomic/) |
| [`Nebius`](https://docs.tokenfactory.nebius.com/quickstart) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-nebius/month)](https://pypi.org/project/langchain-nebius/) |
| [`OpensolrEmbeddings`](https://opensolr.com/langchain) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-opensolr/month)](https://pypi.org/project/langchain-opensolr/) |
| [`EmpirioLabsEmbeddings`](https://docs.empiriolabs.ai) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-empiriolabs/month)](https://pypi.org/project/langchain-empiriolabs/) |
| [`SingleStoreEmbeddings`](https://docs.singlestore.com/cloud/ai/ai-ml-functions/ai-functions/#embed-text) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-singlestore/month)](https://pypi.org/project/langchain-singlestore/) |
| [`Localai`](https://localai.io/features/embeddings/index.html) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-localai/month)](https://pypi.org/project/langchain-localai/) |
| [`PredictionGuardEmbeddings`](https://docs.predictionguard.com/api-reference/api-reference/embeddings) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-predictionguard/month)](https://pypi.org/project/langchain-predictionguard/) |
| [`AimlapiEmbeddings`](https://docs.aimlapi.com/) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-aimlapi/month)](https://pypi.org/project/langchain-aimlapi/) |
| [`DoublewordEmbeddings`](https://docs.doubleword.ai/inference-api/intro-to-doubleword-inference) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-doubleword/month)](https://pypi.org/project/langchain-doubleword/) |
| [`Modelscope`](https://www.modelscope.cn/docs/sdk/pipelines) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-modelscope-integration/month)](https://pypi.org/project/langchain-modelscope-integration/) |
| [`BrainiallEmbeddings`](https://app.brainiall.com/docs) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-brainiall/month)](https://pypi.org/project/langchain-brainiall/) |
| [`PolarDBPGEmbeddings`](https://github.com/polardb/langchain-polardb-pg) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-polardb-pg/month)](https://pypi.org/project/langchain-polardb-pg/) |
| [`ForgeEmbeddings`](https://voxell.ai/forge) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-voxell/month)](https://pypi.org/project/langchain-voxell/) |
| [`TelnyxEmbeddings`](https://developers.telnyx.com/docs/inference/models) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-telnyx/month)](https://pypi.org/project/langchain-telnyx/) |
| [`GreenNodeEmbeddings`](https://greennode.ai/) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-greennode/month)](https://pypi.org/project/langchain-greennode/) |
| [`ANEEmbeddings`](https://github.com/sbryngelson/langchain-aneforge) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-aneforge/month)](https://pypi.org/project/langchain-aneforge/) |
| [`KeiroEmbeddings`](https://github.com/Keirolabs-API/langchain-keiro) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-keiro/month)](https://pypi.org/project/langchain-keiro/) |
| [`Lindorm`](https://help.aliyun.com/en/lindorm/product-overview/product-introduction-overview) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-lindorm-integration/month)](https://pypi.org/project/langchain-lindorm-integration/) |
| [`Netmind`](https://github.com/protagolabs/langchain-netmind) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-netmind/month)](https://pypi.org/project/langchain-netmind/) |
| [`Isaacus`](https://isaacus.com/docs) | [  ![Downloads per month](https://static.pepy.tech/badge/langchain-isaacus/month)](https://pypi.org/project/langchain-isaacus/) |
| [`LlamayEmbeddings`](https://github.com/kosa8ai/langchain-llamay#readme) | N/A |

***

> 官方原文：[src/oss/python/integrations/embeddings/index.mdx](https://github.com/langchain-ai/docs/blob/main/src/oss/python/integrations/embeddings/index.mdx) · 非官方翻译，译文由 AI 生成，以英文原文为准。


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.