llama-index-cli

Warnings

• deprecated The `llama-index-cli` package is officially deprecated and no longer maintained. It will not receive further updates or bug fixes. Users are advised to consider alternatives such as the `create-llama` CLI tool for scaffolding LlamaIndex projects or interacting directly with the core `llama-index` Python library.
  Fix: Migrate to `create-llama` (for project scaffolding) or the main `llama-index` library for programmatic RAG workflows.
• gotcha By default, `llama-index-cli` uses OpenAI for embeddings and LLM, which requires setting the `OPENAI_API_KEY` environment variable. This means that, by default, any local data ingested with this tool will be sent to OpenAI's API.
  Fix: Set the `OPENAI_API_KEY` environment variable or customize the CLI tool to use local models or other providers. Check LlamaIndex documentation for customization options.
• breaking Upgrading the underlying `llama-index` library to versions v0.10.0 and above introduced significant breaking changes, including a massive packaging refactor, deprecation of `ServiceContext`, and changes to import paths. While `llama-index-cli` itself is deprecated, users might encounter issues if their `llama-index` dependency is updated.
  Fix: Refer to the LlamaIndex v0.10 Migration Guide for changes in `llama-index` core. The `llama-index-cli` package included an `upgrade-file` command to assist with these migrations, though its utility is limited now due to the CLI's deprecation.
• gotcha Users installing `llama-index-cli` alongside `chromadb` (a default dependency for RAG) may encounter dependency conflicts, particularly with `onnxruntime` in certain environments (e.g., Conda). This can lead to installation failures.
  Fix: Consider installing `onnxruntime` separately with `SYSTEM_VERSION_COMPAT=0 pip install --no-cache-dir 'onnxruntime>=1.17.1'` before installing `llama-index-cli` and `chromadb`, or use a fresh virtual environment.

Install

• pip install llama-index-cli Latest Version (Deprecated)

Imports

• RagCLI

  from llama_index.cli.rag_cli import RagCLI

  While this was the programmatic entry for customizing the RAG CLI, the `llama-index-cli` package is deprecated. Direct usage is discouraged in favor of the main `llama-index` library or its modern CLI tool, `create-llama`.

Quickstart

This quickstart demonstrates how to use the `llamaindex-cli` tool to ingest a local file and query it using Retrieval-Augmented Generation (RAG). It requires an OpenAI API key by default. Note that `llama-index-cli` is deprecated; for new projects, consider using `create-llama` or directly integrating with the `llama-index` Python library.

import os
import subprocess

# NOTE: llama-index-cli is deprecated. Use the main llama-index library
# or the 'create-llama' CLI for new projects.

# 1. Set your OpenAI API key (required by default for embeddings/LLM)
# Replace 'YOUR_OPENAI_API_KEY' with your actual key or set it as an env var.
os.environ['OPENAI_API_KEY'] = os.environ.get('OPENAI_API_KEY', 'YOUR_OPENAI_API_KEY')

# Create a dummy file for demonstration
with open('example.txt', 'w') as f:
    f.write('LlamaIndex is a data framework for LLM applications. It helps connect custom data sources with LLMs.')

print("Ingesting 'example.txt' and asking a question via llamaindex-cli...")

# 2. Ingest a file and ask a question
try:
    result = subprocess.run(
        ['llamaindex-cli', 'rag', '--files', 'example.txt', '--question', 'What is LlamaIndex?'],
        capture_output=True, text=True, check=True
    )
    print("--- CLI Output ---")
    print(result.stdout)
    if result.stderr:
        print("--- CLI Errors (if any) ---")
        print(result.stderr)
except FileNotFoundError:
    print("Error: 'llamaindex-cli' command not found. Ensure the package is installed and in your PATH.")
except subprocess.CalledProcessError as e:
    print(f"Error running command: {e}")
    print(f"Stdout: {e.stdout}")
    print(f"Stderr: {e.stderr}")

# Clean up the dummy file
os.remove('example.txt')