As with other packages, we provide a simple and intuitive installment via pypi.

pip install argilla

Package extras#

Our Python package requires some extras that might be downloaded to facilitate more custom workflows.

  • pip install "argilla[listeners]": the argilla.listeners-module allows for the usage of background processes to monitor dataset changes and schedule jobs.

  • pip install "argilla[server]": the Argilla FastAPI server can be deployed locally to test development changes or custom configs.

  • pip install "argilla[postgresql]": the default data management is done with built-in sqlite but can be replaced with a PostgreSQL database.

  • pip install "argilla[integrations]": integrations with other libraries/frameworks are available to use.

  • pip install "argilla[tests]" When running tests as a developer, you might need third-party integration packages to test end-to-end workflows.

Install from develop#

If you want the cutting-edge version of Argilla with the latest changes and experimental features, follow the steps below in your terminal. Be aware that this version might be unstable!

First, you need to install the develop version of our Python client:

 pip install -U git+

Then, the easiest way to get the develop version of our web app up and running is via docker-compose:


For now, we only provide the develop version of our web app via docker. If you want to run the web app of the develop branch without docker, we refer you to our development setup.

 # get the docker-compose yaml file
 mkdir argilla && cd argilla
 wget -O docker-compose.yaml
 # use the `develop` image of the argilla container instead of the latest
 sed -i 's/argilla:latest/argilla:develop/' docker-compose.yaml
 # start all services
 docker-compose up

If you want to use vanilla docker (and have your own Elasticsearch instance running), you can just use our develop image:

docker run -p 6900:6900 -e "ARGILLA_ELASTICSEARCH=<your-elasticsearch-endpoint>" --network argilla-net --name argilla argilla/argilla-server:develop