How to beginner · 4 min read

How to load multiple files with LlamaIndex

Q: How to load multiple files with LlamaIndex

Use LlamaIndex's SimpleDirectoryReader or Document loaders to load multiple files by specifying a directory or iterating over file paths. Then create an index with GPTVectorStoreIndex or other index classes to process all loaded documents together.

Quick answer

Use LlamaIndex's SimpleDirectoryReader or Document loaders to load multiple files by specifying a directory or iterating over file paths. Then create an index with GPTVectorStoreIndex or other index classes to process all loaded documents together.

PREREQUISITES

Python 3.8+
pip install llama-index>=0.6.0
OpenAI API key (free tier works)
Set environment variable OPENAI_API_KEY

Setup

Install llama-index via pip and set your OpenAI API key in the environment.

bash

pip install llama-index>=0.6.0

Step by step

This example loads multiple text files from a directory using SimpleDirectoryReader, then creates a vector index with GPTVectorStoreIndex for querying.

python

import os
from llama_index import GPTVectorStoreIndex, SimpleDirectoryReader

# Ensure your OpenAI API key is set in environment
# export OPENAI_API_KEY=os.environ["OPENAI_API_KEY"]

# Load all documents from a directory
loader = SimpleDirectoryReader('data')  # 'data' folder with multiple text files
documents = loader.load_data()

# Create an index from the loaded documents
index = GPTVectorStoreIndex.from_documents(documents)

# Query the index
query = "What is the main topic of these documents?"
response = index.query(query)
print(response.response)

output

The main topic of these documents is ... (depends on your files)

Common variations

Use Document class to load files individually and combine into a list.
Load PDFs or other formats with specialized loaders like PDFReader.
Use async methods if supported for large datasets.
Switch index types, e.g., GPTListIndex or GPTTreeIndex, depending on use case.

python

from llama_index import Document

file_paths = ['data/file1.txt', 'data/file2.txt']
documents = []
for path in file_paths:
    with open(path, 'r', encoding='utf-8') as f:
        text = f.read()
        documents.append(Document(text=text, doc_id=path))

index = GPTVectorStoreIndex.from_documents(documents)
response = index.query("Summarize the content.")
print(response.response)

output

Summary of the content ...

Troubleshooting

If you see FileNotFoundError, verify your directory path and file names.
For encoding errors, ensure files are UTF-8 encoded or specify encoding explicitly.
If the index creation fails, check your OpenAI API key and network connectivity.

✅

Key Takeaways

Use SimpleDirectoryReader to load multiple files from a folder easily.
Combine loaded documents into a list and create an index with GPTVectorStoreIndex.
You can load files individually with Document for more control.
Check file paths and encoding to avoid common loading errors.
Choose the index type based on your querying needs.

Verified 2026-04 · gpt-4o

Verify ↗