Lexy is a data platform for building AI-powered applications. You can use Lexy to:
- Build Retrieval Augmented Generation (RAG) applications.
- Provide relevant context for AI agents.
- Add long-term memory to chat-based applications.
- Extract structured data from unstructured documents.
To run Lexy locally, you'll need Docker installed. You'll also need Python 3.11 or greater. To install Lexy, you can either build from source (recommended) or install from PyPI (standalone mode).
git clone https://github.com/lexy-ai/lexy.gitFirst create a virtual environment. Make sure that you're using Python 3.11 or greater. You can check your Python
version by running python3 --version.
# Change to the lexy directory
cd lexy
# Create a virtualenv
python3 -m venv venv
source venv/bin/activateNext, install the dev dependencies. This will also create a .env file in the working directory if it doesn't
exist already.
# Install dev dependencies
make install-devTo build and run the docker containers, run the following command:
# Build and run docker containers
docker compose up --build -dTo stop the containers, run docker compose down.
Use this method if you're already familiar with Lexy and are using it as a dependency in your own project. You'll need to provide instances of the Lexy server, worker, and database. See the Quickstart guide for more information and a detailed example.
# Install server
pip install lexy[all]
# Install Python client
pip install lexy-py
# Initialize Lexy
lexy init
# Optional: Create a docker-compose file
lexy dockerIn order to upload and store files to Lexy, you'll need to configure either Amazon S3 or Google Cloud Storage.
To configure Amazon S3, you can use aws configure on the command line (recommended) or set the value of
AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY in your .env file.
If you choose to use aws configure, make sure to uncomment the following line in docker-compose.yaml
in order for your credentials to be available inside of the lexyserver container.
volumes:
# Uncomment the following line to mount local AWS credentials
- $HOME/.aws/credentials:/root/.aws/credentials:roYou'll also need to specify an S3 bucket for file storage (for which your AWS credentials should have full access).
You can do so by adding DEFAULT_STORAGE_BUCKET=<name-of-your-S3-bucket> to your .env file. Remember to rebuild
your dev containers for the change to take effect (run make rebuild-dev-containers on the command line).
# Example .env settings for Amazon S3
DEFAULT_STORAGE_SERVICE=s3
DEFAULT_STORAGE_BUCKET=your_s3_bucket_name
AWS_ACCESS_KEY_ID=your_aws_access_key_id
AWS_SECRET_ACCESS_KEY=your_aws_secret_access_keyTo use Google Cloud Storage, you'll need to set DEFAULT_STORAGE_SERVICE=gcs in your .env file. You'll also need to
create a service account, and set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path of your service
account credentials. You can do so by adding the following to your .env file:
GOOGLE_APPLICATION_CREDENTIALS=<path-to-your-service-account-credentials>You'll also need to specify a GCS bucket for file storage (for which your service account should have full access).
You can do so by adding DEFAULT_STORAGE_BUCKET=<name-of-your-GCS-bucket> to your .env file. Remember to rebuild
your dev containers for the change to take effect (run make rebuild-dev-containers on the command line).
# Example .env settings for Google Cloud Storage
DEFAULT_STORAGE_SERVICE=gcs
DEFAULT_STORAGE_BUCKET=your_gcs_bucket_name
GOOGLE_APPLICATION_CREDENTIALS=/path/to/your/service-account-credentials.jsonTo use OpenAI embeddings in Lexy, you'll need to set the OPENAI_API_KEY environment variable. You can do so by adding
the following to your .env file:
OPENAI_API_KEY=<your-openai-api-key>Do this before building your docker containers. Or, if you've already run docker compose up, you can run the
following to rebuild the server and worker containers.
# Rebuild the server and worker containers
make rebuild-dev-containersThe Lexy server is a RESTful API that provides endpoints for storing and retrieving documents, applying transformations, and managing collections and indexes.
The API is documented using Swagger. You can view the Swagger UI in the
REST API docs or access it locally at
http://localhost:9900/docs when running the Lexy server.
lexy-py is the Python SDK used to interact with the Lexy server.
The SDK provides a LexyClient class that you can use to interact with the Lexy server. Here's an example of how to list
collections using the Python SDK:
from lexy_py import LexyClient
lx = LexyClient()
lx.list_collections()For more information on how to use the Python SDK, see the Python SDK reference.
Lexy comes with a built-in dashboard to visualize pipelines. This is still under development, but you can run it locally. See the dashboard README for more details.
To start the dashboard, make sure you have Node.js installed. Then, from the root directory, run the following commands:
cd dashboard
npm install
npm run devThe dashboard will be running at http://localhost:3000.
The server will be running at http://localhost:9900. In addition, you can find the following services.
| Service | URL | Notes |
|---|---|---|
| Lexy API | http://localhost:9900/docs | Swagger API docs |
| Flower | http://localhost:5556 | Celery task monitor |
| RabbitMQ | http://localhost:15672 | Username: guest, Password: guest |
| Postgres | http://localhost:5432 | Database: lexy, Username: postgres, Password: postgres |
| Project docs | http://localhost:8000 | Run make serve-docsUsername: lexy, Password: guest |
| Lexy dashboard | http://localhost:3000 | Dashboard to show pipelines (WIP) |
If your virtualenv keeps getting bjorked by PyCharm, make sure that you're following the instructions above verbatim,
and using venv instead of .venv for the path of your virtual environment.
If you encounter any issues while installing or using Lexy, please open an issue on GitHub.

