Kaggle

What it does

Kaggle Notebooks takes in Python or R code, written as Jupyter notebook cells, scripts or RMarkdown, plus data sources attached from Kaggle's datasets, competitions and models. It runs that code in a hosted environment and produces versioned outputs: console results, output files, visualizations and a saved record of the environment used.

People use it to explore a dataset, follow a tutorial, build an analysis, or prepare competition submissions. Each notebook version is tied to a Docker image with preinstalled packages, and extra packages can be added with pip, or with a Dependency Manager for internet-disabled competitions. Distinct features include a large library of public notebooks you can search, filter by tag and fork, optional accelerators, scheduled runs, and connections to Google BigQuery, Cloud Storage and AutoML.

It suits students, analysts and machine learning practitioners who want a browser-based workspace without local setup. A Kaggle account is required to create notebooks. A TPU v3-8 can be added for free, but availability is limited and queues are possible. The BigQuery integration needs a billing-enabled Google Cloud account for non-public data, and Cloud Storage and AutoML are paid. The environment is code-focused rather than a no-code tool.

What it can do

• Write and run Jupyter notebooks in Python or R
• Run scripts and RMarkdown scripts, including selected lines
• Attach datasets and models as data sources
• Install extra packages via pip or a Dependency Manager for offline installs
• Schedule notebooks to run automatically
• Search and fork public community notebooks by tag, votes or recency

Works with

• Google BigQuery
• Google Cloud Storage
• Google AutoML

Example workflows

Learn from a public notebook

Input: a topic tag or dataset name. Steps: search notebooks or use tag:classification, open a popular notebook on the dataset's Notebooks tab, fork it, then edit and rerun cells. Output: your own versioned copy with modified code and results.

Prepare a competition submission

Input: a competition dataset and your model code. Steps: create a script or notebook, attach the data, add dependencies through the Dependency Manager if internet is disabled, commit a version. Output: a committed version with output files for submission.

Before you choose

• Free TPU availability is limited and may involve queues at busy times.
• BigQuery, Cloud Storage and AutoML use can incur charges on a linked Google Cloud billing account.
• Competitions with internet disabled require dependencies set up through the Dependency Manager.

Listed under

• Data
• AI benchmarks
• Community
• Competitions
• Datasets
• Machine learning
• Model evaluation

Get started

Create a Kaggle account, then choose Create > New Notebook.

1. Sign up for a Kaggle account
2. Click Create > New Notebook and choose Python or R
3. Add data sources in the Input tab and set session options such as accelerator

Kaggle provides limited accelerator availability, including free TPU notebook access. Linked Google Cloud services can incur charges; check notebook quotas and billing settings.

We couldn’t find published prices for Kaggle. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 28 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Getting Started on Kaggle | Kaggle ↗www.kaggle.com

Company website

kaggle.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 28 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.