Installation

AdalFlow is available in Python.

1. Install AdalFlow

To install the package, run:

pip install adalflow

If you know you will need openai and faiss-cpu, you can do so with:

pip install adalflow[openai, faiss-cpu]

Note

Check the Optional Packages section for more information on the available packages.

2. Set up API keys

A .env file is recommended. You can have it at your project root directory. Here is an example:

OPENAI_API_KEY=YOUR_API_KEY_IF_YOU_USE_OPENAI
GROQ_API_KEY=YOUR_API_KEY_IF_YOU_USE_GROQ
ANTHROPIC_API_KEY=YOUR_API_KEY_IF_YOU_USE_ANTHROPIC
GOOGLE_API_KEY=YOUR_API_KEY_IF_YOU_USE_GOOGLE
COHERE_API_KEY=YOUR_API_KEY_IF_YOU_USE_COHERE
HF_TOKEN=YOUR_API_KEY_IF_YOU_USE_HF

3. Load environment variables

You can use the following import:

from adalflow.utils import setup_env

setup_env()

Or, you can load it yourself with python-dotenv:

from dotenv import load_dotenv
load_dotenv()  # This loads the environment variables from `.env`.

This setup ensures that AdalFlow can access all necessary configurations during runtime.

4. Install Optional Packages

AdalFlow currently has built-in support for (1) OpenAI, Groq, Anthropic, Google, and Cohere, and (2) FAISS and Transformers. You can find all optional packages at OptionalPackages. Make sure to install the necessary SDKs for the components you plan to use. Here is the list of our tested versions:

openai = "^1.12.0"
groq = "^0.5.0"
faiss-cpu = "^1.8.0"
sqlalchemy = "^2.0.30"
pgvector = "^0.3.1"
torch = "^2.3.1"
anthropic = "^0.31.1"
google-generativeai = "^0.7.2"
cohere = "^5.5.8"

You can install the optional packages with either pip install package_name or pip install lightrag[package_name].