Overview

Main package aggregating LlamaIndex.TS functionality

Overview

The llamaindex package is the primary entry point for LlamaIndex.TS. It aggregates and re-exports functionality from @llamaindex/core and other packages, providing a unified API for building LLM applications.

Installation

npm install llamaindex

Package Structure

The package provides multiple runtime-specific entry points:

• Node.js: Default import with file system support
• Edge Runtime: Vercel Edge, Cloudflare Workers
• React Server Components: Next.js App Router
• Cloudflare Workers: Workerd-specific optimizations

Sub-module Exports

The package uses tree-shakeable sub-module exports:

import { VectorStoreIndex } from "llamaindex";
import { OpenAI } from "@llamaindex/openai";
import { SimpleChatEngine } from "llamaindex/engines/chat";
import { RetrieverQueryEngine } from "llamaindex/engines/query";
import { SentenceSplitter } from "llamaindex/node-parser";

Core Modules

Indices

import {
  VectorStoreIndex,
  SummaryIndex,
  KeywordTableIndex
} from "llamaindex";

Engines

import { RetrieverQueryEngine } from "llamaindex/engines/query";
import { ContextChatEngine, SimpleChatEngine } from "llamaindex/engines/chat";

Storage

import {
  SimpleVectorStore,
  StorageContext,
  SimpleDocumentStore,
  SimpleIndexStore
} from "llamaindex/storage";

Node Parsing

import { SentenceSplitter, MarkdownNodeParser } from "llamaindex/node-parser";

Ingestion

import { IngestionPipeline } from "llamaindex/ingestion";

Global Settings

import { Settings } from "llamaindex";
import { OpenAI } from "@llamaindex/openai";
import { OpenAIEmbedding } from "@llamaindex/openai";

// Configure default LLM
Settings.llm = new OpenAI({ model: "gpt-4" });

// Configure default embedding model
Settings.embedModel = new OpenAIEmbedding({ model: "text-embedding-3-small" });

// Configure chunk size
Settings.chunkSize = 512;
Settings.chunkOverlap = 50;

Quick Start Example

import { VectorStoreIndex, Settings, Document } from "llamaindex";
import { OpenAI, OpenAIEmbedding } from "@llamaindex/openai";

// Configure settings
Settings.llm = new OpenAI({ model: "gpt-4" });
Settings.embedModel = new OpenAIEmbedding();

// Create documents
const documents = [
  new Document({ text: "LlamaIndex is a data framework for LLM applications." }),
  new Document({ text: "It provides tools for data ingestion, indexing, and retrieval." })
];

// Build index
const index = await VectorStoreIndex.fromDocuments(documents);

// Query
const queryEngine = index.asQueryEngine();
const response = await queryEngine.query({
  query: "What is LlamaIndex?"
});

console.log(response.response);

Runtime Compatibility

Node.js

import { VectorStoreIndex } from "llamaindex";
import fs from "fs";

// Full Node.js API support
const documents = await SimpleDirectoryReader.loadData("./docs");

Edge Runtime

import { VectorStoreIndex } from "llamaindex/edge";

// No file system access
// Use fetch or other edge-compatible APIs

Cloudflare Workers

import { VectorStoreIndex } from "llamaindex/workerd";

// Optimized for Cloudflare Workers runtime

Environment Variables

Common environment variables:

# OpenAI
OPENAI_API_KEY=sk-...

# Anthropic
ANTHROPIC_API_KEY=sk-ant-...

# Vector Stores
PINECONE_API_KEY=...
PINECONE_INDEX_NAME=...

TypeScript Support

Full TypeScript support with type inference:

import { VectorStoreIndex, Document } from "llamaindex";
import type { NodeWithScore } from "@llamaindex/core/schema";

const index = await VectorStoreIndex.fromDocuments(documents);
const nodes: NodeWithScore[] = await index.asRetriever().retrieve("query");

Next.js Integration

// app/api/chat/route.ts
import { VectorStoreIndex, Settings } from "llamaindex";
import { OpenAI } from "@llamaindex/openai";

Settings.llm = new OpenAI({ model: "gpt-4" });

export async function POST(request: Request) {
  const { message } = await request.json();
  
  const index = await VectorStoreIndex.fromDocuments(documents);
  const chatEngine = index.asChatEngine();
  
  const response = await chatEngine.chat({ message });
  
  return Response.json({ response: response.response });
}

Deprecated Features

The following features are deprecated:

• Agents (llamaindex/agent): Use @llamaindex/workflow instead
• ReACTAgent: Migrate to workflow-based agents

See Also

• Agents
• Engines
• Indices
• Ingestion