0
# Embeddings
1

2
Text embedding generation and similarity operations for semantic search and retrieval in LlamaIndex.TS.
3

4
## Import
5

6
```typescript
7
import { OpenAIEmbedding, Settings } from "llamaindex";
8
```
9

10
## Overview
11

12
Embeddings convert text into numerical vector representations that capture semantic meaning, enabling similarity search and retrieval operations in LlamaIndex.TS.
13

14
## Base Embedding Interface
15

16
```typescript { .api }
17
interface BaseEmbedding {
18
  getTextEmbedding(text: string): Promise<number[]>;
19
  getQueryEmbedding(query: string): Promise<number[]>;
20
  
21
  embedBatchSize?: number;
22
  dimensions?: number;
23
}
24
```
25

26
## OpenAI Embeddings
27

28
```typescript { .api }
29
class OpenAIEmbedding implements BaseEmbedding {
30
  constructor(options?: {
31
    model?: string;
32
    dimensions?: number;
33
    apiKey?: string;
34
  });
35
  
36
  getTextEmbedding(text: string): Promise<number[]>;
37
  getQueryEmbedding(query: string): Promise<number[]>;
38
}
39
```
40

41
## Basic Usage
42

43
```typescript
44
import { OpenAIEmbedding, Settings } from "llamaindex";
45

46
// Configure global embedding model
47
Settings.embedModel = new OpenAIEmbedding({
48
  model: "text-embedding-3-large",
49
  dimensions: 1536,
50
});
51

52
// Generate embeddings
53
const text = "LlamaIndex is a data framework";
54
const embedding = await Settings.embedModel.getTextEmbedding(text);
55
console.log("Embedding dimensions:", embedding.length);
56
```
57

58
## Similarity Functions
59

60
```typescript { .api }
61
function similarity(embedding1: number[], embedding2: number[]): number;
62

63
function getTopKEmbeddings(
64
  queryEmbedding: number[],
65
  embeddings: number[][],
66
  k: number
67
): number[];
68
```
69

70
## Usage with Indices
71

72
```typescript
73
// Embeddings are automatically used by VectorStoreIndex
74
const index = await VectorStoreIndex.fromDocuments(documents);
75
// Uses Settings.embedModel to generate embeddings for documents
76
```