Install packages once.
Use them everywhere.
The pnpm of Python. Drop-in replacement for pip and uv that saves GBs of disk space using shared global symlinks.
Stop wasting your disk space on duplicate dependencies.
Traditional pip installs complete package copies into every virtual environment. Developing 5 projects with Pandas? That's 5 copies of the same binary blobs eating your SSD.
The Symlink Revolution
Pepip stores every version of every package in a single, central store on your machine. Local environments simply link to these files.
Global Store
One immutable instance of every package at
~/.pepip/store
Content Addressing
Hard-links and symlinks prevent accidental mutations and maximize sharing.
Zero Copy
Creating an environment is as fast as your OS can create directory entries.
Lightning Fast (Powered by uv)
Pepip uses the high-performance 'uv' resolver and installer under the hood. It’s written in Rust, but optimized for global storage.
pepip install numpy scipy torch
| Resolved in 120ms
Disk Efficient
Save up to 90% of disk space in large monorepos or multi-project workstations.
Team Shared
Set PEPIP_HOME to a shared drive to sync dependencies
across your entire internal network instanty.
Drop-in Replacement
Use the commands you already know. No new learning curve, no specialized configuration files required.
Numbers don't lie.
We benchmarked pepip against standard pip and vanilla uv on a machine with 50 existing virtual environments. The difference isn't just incremental—it's transformative.
Storage Efficiency
Get started in seconds.
pip install pepip
pepip install numpy scipy scikit-learn
Works on macOS, Linux, and Windows.