Polymarket Open Data

The open release ships a recorded snapshot — 897,192 L2 snapshots across 805 resolved crypto markets on the same 250ms grid the full archive uses — under CC BY 4.0 with DataCite registration and a versioned DOI (10.5281/zenodo.22084114).

Open here means the artifact exists for anyone: no key, no paywall, no plan gate, and the license allows sharing and adaptation with attribution.

CC BY 4.0 permits copying, redistribution, and derivative use in any medium for any purpose (including commercial), conditioned on attribution. It is the same family as the license declared on the dataset record, and the honest reading applies it to the artifact, not to proprietary captures.

The boundary: the open release is a scientific artifact; the full rolling archive (all markets, 250ms, live capture, licensing) is the commercial service. Open and commercial coexist by separation of files, not by ambiguity.

The release carries prices, books, metrics, and outcomes on the shared grid, so a teaching model or a paper's reproducibility section can stand on the sample without a key. Its fidelity is the archive's own: 250ms cadence, flags intact, outcomes attached.

The honest limits: a snapshot in time, 805 markets, not the live stream — the open artifact proves the method and the archive serves the motion.

Open data is a trust artifact and a proof of work, not a business model: the site publishes it because a citable, re-verifiable release is how datasets earn credibility — and the citation guide is the other half of that promise.

• Citation guide
• Data quality
• Datasets guide