AI Potluck
Infrastructure / Storage & Retrieval

Oxen

Oxen.ai

Oxen is a data version control system built for large repositories of files, from training data and model weights to game assets. It mirrors git's workflow (init, add, commit, push, pull) while handling millions of files and terabytes of data, with a Rust core (`liboxen`), a CLI, Python bindings, and a self-hostable Oxen Server for remotes.

oxenai on PyPI is declared as the measurement identity. It is built from this repository's `oxen-python` directory and compiles liboxen into the wheel, so installing it gives you Oxen (init, add, commit, push, pull run locally) rather than a client to an Oxen Server, and its metadata points back at Oxen-AI/Oxen. It is a minority channel next to the CLI, so the adoption note treats the count as a lower bound. Verified 2026-09-02 via the GitHub API, the PyPI metadata and the LICENSE body.

Openness

5 high confidence
5.0
license
Apache-2.0(OSI)
source
public
core-gated
ungated

Apache-2.0 license body confirmed. The repository is public and unarchived and builds the CLI, the Rust core and the Oxen Server, and the README describes no paid tier, enterprise edition or license-gated build beside it - Oxen Hub is a separately sold hosted service rather than a piece withheld from the published code.

Adoption

1 medium confidence
1.0

4,302 PyPI downloads of `oxenai` in the trailing 30 days, which lands in the <10K band of the software usage scale, level 1. oxenai is built from this repository's `oxen-python` directory and compiles liboxen into the wheel, so installing it gives you Oxen itself rather than a client to an Oxen Server, and it is the one declared artifact with a countable channel. It undercounts - the CLI is installed through Homebrew, cargo and a shell script that publish no figures - which is why confidence is medium rather than high. The stars scale would read one level higher (1,183 stars, 1K-10K), but a declared package artifact is the measurement identity and its count is what the warehouse routes to.

Capability

3 medium confidence
3.0

Banded on the category feature matrix as working single-purpose store or index, level with lakefs - a versioned data layer rather than a distributed, cluster-capable system, with no indexing or retrieval beyond file and diff lookups.

Verified 2026-09-02