# Tine

> Tine is an AI native workspace for branching notebooks, reproducible data science experiments, and agent accessible analytics workflows.

Tine provides a branching notebook runtime for AI agents and humans. It is designed for data science, analytics, visualization, modelling, reproducible experiments, and workflows where agents need structured access to notebook state.

Tine Cloud is the hosted collaboration experience for shared projects, teams, and AI agents. Tine Local is available today for local notebook runtime workflows.

## Product

- [Homepage](https://tine.sh/): Product overview, local install options, and Tine Cloud waitlist.
- [GitHub repository](https://github.com/tinelabs/tine): Source code, releases, issues, and project activity.
- [Tine Cloud waitlist](https://tally.so/r/Y5gzEz): Request early access to hosted collaboration.

## Install

- [Python package](https://pypi.org/project/tine/): Install with `pip install tine`, then run `tine serve --open`.
- [macOS Apple Silicon installer](https://github.com/tinelabs/tine/releases/latest/download/tine-macos-aarch64.dmg): Local desktop installer for macOS.
- [Windows x64 installer](https://github.com/tinelabs/tine/releases/latest/download/tine-windows-x86_64.msi): Local desktop installer for Windows.

## Topics

- Branching notebooks
- AI native workspace
- Reproducible data science experiments
- Analytics and visualization workflows
- MCP/API agent access
- Local notebook runtime
- Hosted cloud collaboration

## Optional

- [Sitemap](https://tine.sh/sitemap.xml): Indexable public URLs.
- [Robots policy](https://tine.sh/robots.txt): Crawler access policy.