Datasets

Keep the datasets you work with in one place, add new ones from Kaggle, and download their files straight into your workspace so you can query, profile, or train on them.

What you can do

• Browse your datasets — every dataset shows up as a card with its name, source, size, file types, and whether its files are downloaded yet.
• Add from Kaggle — search Kaggle and import a dataset into your catalog.
• Download files — pull a dataset's files into your workspace with one click; progress streams into the Terminal panel, and the card then points at the downloaded folder.
• Reveal — open a downloaded dataset's folder in the file explorer.
• Remove — take a dataset out of your catalog (downloaded files are left in place).

Once files are in your workspace, open them with DuckDB or the Data Profiler.

Requirements

• Add from Kaggle needs the Kaggle Datasets extension.
• Downloading Kaggle files uses the Kaggle CLI: install it with pip install kaggle and add your API token at ~/.kaggle/kaggle.json (Kaggle → Account → Create New API Token). Managed cloud workspaces download server-side, with no local setup.

Getting started

Open the Datasets sidebar from the activity bar and press + to add your first dataset.