---
id: jupyter-cache
version: "1.0.1"
license: unclear
license_treatment: permissive
maintenance: active
---

# jupyter-cache — A defined interface for working with a cache of jupyter notebooks.

License: permissive · Maintenance: active · Downloads: 1.0M/mo

## What it is and what it does

jupyter-cache provides a persistent cache layer for Jupyter notebook execution outputs. It stores the outputs of notebooks so that when you rebuild documentation or re-run notebooks, the cached results can be reused instead of re-executing the entire notebook—useful when notebooks contain long-running computations or generate outputs in formats that don't naturally store results.

The package is designed for notebooks with deterministic execution: same environment, no random code, no time-dependent external resources. It separates content edits from code cell changes, so only actual code modifications trigger re-execution. The package is actively used by jupyter-book to enable fast document rebuilds and is maintained by the Executable Books project.

Use it for:

- Speed up documentation builds in jupyter-book by caching notebook execution outputs across rebuilds.
- Avoid re-running expensive computations (long simulations, data processing) when only notebook text or formatting changes.
- Maintain consistent outputs across multiple notebook runs in CI/CD pipelines without re-executing deterministic code.
- Cache outputs from notebooks that depend on external data or scripts that rarely change.
- Enable parallel notebook execution with shared cache for teams working on the same documentation.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Caches Jupyter notebook execution outputs to avoid re-running notebooks with deterministic outputs, enabling fast rebuilds of documentation and notebooks that depend on long-running computations.

Yes, if you maintain Jupyter notebooks or documentation that re-builds frequently and contains long-running code. The package is actively maintained, has no known vulnerabilities, and low install friction. Not necessary for one-off notebooks or those with non-deterministic outputs.

## Install

pip install jupyter-cache
uv add jupyter-cache
poetry add jupyter-cache

## Installing jupyter-cache

Before you install: Low friction install with 8 straightforward runtime dependencies. Actively maintained with recent commits; last release was 637 days ago but the repository shows ongoing activity as of 2026-08-03.

License in practice: Licensed under MIT (permissive), so you can use, modify, and distribute the package freely with minimal restrictions.

Quickstart:

pip install jupyter-cache

from jupyter_cache import CacheManager
cm = CacheManager()
cm.cache_notebook_file('notebook.ipynb')

Requires Python 3.9 or later; notebooks must have deterministic outputs (same environment, no random code, no time-dependent external resources).

Verify before relying:

- Whether the cache correctly handles notebook cell dependencies and invalidation across Python package updates or git branch changes.
- Performance characteristics and cache size limits for large numbers of cached notebooks.
- Compatibility with recent versions of jupyter-book and other downstream tools that depend on this package.

## Package facts

- License: not declared (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 1.0M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

jupyter notebook caching, cache notebook execution, avoid re-running notebooks, notebook output cache, deterministic notebook execution, fast notebook rebuilds, jupyter-book caching, notebook-caching, documentation-build, jupyter-ecosystem

[View on SkillFed](https://skillfed.io/packages/jupyter-cache) · [View on PyPI](https://pypi.org/project/jupyter-cache/)