How to beginner · 3 min read

How to use OpenAI embeddings in LlamaIndex

Q: How to use OpenAI embeddings in LlamaIndex

Use the OpenAIEmbeddings class from langchain_openai with your OpenAI API key, then pass it to LlamaIndex's ServiceContext to build an index with embedded vectors. This enables semantic search and retrieval using OpenAI's embedding models.

Quick answer

Use the OpenAIEmbeddings class from langchain_openai with your OpenAI API key, then pass it to LlamaIndex's ServiceContext to build an index with embedded vectors. This enables semantic search and retrieval using OpenAI's embedding models.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install llama-index langchain_openai openai>=1.0

Setup

Install the required packages and set your OpenAI API key as an environment variable.

Install packages: pip install llama-index langchain_openai openai
Set environment variable in your shell: export OPENAI_API_KEY='your_api_key' (Linux/macOS) or setx OPENAI_API_KEY "your_api_key" (Windows)

bash

pip install llama-index langchain_openai openai

Step by step

This example shows how to create an embedding model using OpenAI, build a LlamaIndex vector index from documents, and query it.

python

import os
from llama_index import SimpleDirectoryReader, ServiceContext, VectorStoreIndex
from langchain_openai import OpenAIEmbeddings

# Initialize OpenAI embeddings with API key from environment
embeddings = OpenAIEmbeddings(openai_api_key=os.environ["OPENAI_API_KEY"])

# Create a ServiceContext with the embeddings
service_context = ServiceContext.from_defaults(embed_model=embeddings)

# Load documents from a directory (replace 'docs' with your folder)
docs = SimpleDirectoryReader('docs').load_data()

# Build the vector index with the documents and embeddings
index = VectorStoreIndex.from_documents(docs, service_context=service_context)

# Query the index
query = "What is LlamaIndex?"
response = index.query(query)
print(response.response)

output

LlamaIndex is a library that helps you build vector indexes over your documents for semantic search and retrieval.

Common variations

Use different embedding models by passing other embed_model implementations to ServiceContext.
Use GPTVectorStoreIndex instead of VectorStoreIndex for older versions.
For async usage, wrap calls with async functions and use async-compatible clients.

Troubleshooting

If you get authentication errors, verify your OPENAI_API_KEY environment variable is set correctly.
If embeddings fail, check your network connection and API usage limits.
Ensure document paths are correct to avoid file not found errors.

✅

Key Takeaways

Use OpenAIEmbeddings from langchain_openai to generate embeddings for LlamaIndex.
Pass the embeddings to ServiceContext to integrate with LlamaIndex vector indexes.
Load your documents, build the index, then query it for semantic search capabilities.

Verified 2026-04 · gpt-4o, OpenAIEmbeddings

Verify ↗