Quickstart

Install the library

The library requires Python 3.11 or later.

pip install -U gwenlake

uv add gwenlake

To try the latest development version:

pip install -U git+https://github.com/gwenlake/gwenlake-python

Set your API key

Create an API key in the platform, then export it:

export GWENLAKE_API_KEY='sk-...'

See Authentication for credential files, profiles and OAuth2 clients.

Make your first call

from gwenlake import Gwenlake

client = Gwenlake()

for d in client.datasets.list():
    print(d["slug"], d["id"])

The same listing is one command away in a terminal:

gwenlake datasets

Query a dataset

SQL statements run on DuckDB. A dataset is referenced by its handle, '<project_slug>.<dataset_slug>':

result = client.statements.create(
    statement="SELECT * FROM 'flights.flight-data' LIMIT 10",
    format="json",
)
for row in result["data"]:
    print(row)

Write your first transform

A transform reads datasets, computes, and writes a dataset. The platform records the lineage.

from gwenlake.transforms import transform_df, Input, Output

@transform_df(
    flights=Input("flights.flight-data"),
    delayed=Output("flights.delayed"),
)
def keep_delayed(flights):
    return flights[flights["arr_delay"] > 15]

keep_delayed(client)

Next steps

Core concepts

How projects, datasets, connections and models fit together.

Transforms

Everything the decorators can do: files, several outputs, append mode.

Chat and embeddings

Call the inference models with the same client.

Examples

Runnable scripts in the repository.

Get started

Concepts

Workspace

Data

Code

Build

Models

Inference

Agents

Client

Changelog

Next steps

Core concepts

Transforms

Chat and embeddings

Examples

​Next steps

Core concepts

Transforms

Chat and embeddings

Examples

Next steps