Public data,
ready to query.

Selected open datasets — company registries, SEC filers, industry codes — searchable and queryable in your browser, and yours to take anywhere with one line of SQL. No signup. No install.

Explore datasets Request a dataset →

3 datasets3.37M+ recordsopen licenses (CC0 / public domain)sources & provenance →

Find it fast

Full-text search across millions of typed records — 3.36M legal entities answer in a couple of seconds, with column profiles built in.

Query it here

A real SQL grid in your browser: filter, group, aggregate, export CSV. The engine runs client-side — your queries never touch a server.

Take it anywhere

One ATTACH line exposes every dataset in DuckDB, Python or R. Stable URLs, open licenses, provenance and as-of dates included.

INSTALL ducklake; INSTALL httpfs;
ATTACH 'ducklake:https://openlake.meridian.online/catalog/open.ducklake' AS meridian (READ_ONLY);
SELECT * FROM meridian.gleif LIMIT 10;

What should we add next?

The commons grows by request. Tell us the public dataset you need — requests directly decide what ships next.

Coming next

Meridian for Excel & Sheets

Look up entities and codes from inside your spreadsheet — =MERIDIAN_LEI("Tesla"). Join the early-access list.

The maintained crosswalk

LEI ↔ SEC CIK ↔ ticker ↔ NAICS, resolved and refreshed — the join you shouldn't have to rebuild. Early access for analysts and teams.

How every dataset is made

The datasets and the tools are one project: type every column, verify every relationship, make every pipeline reproducible — open source, so you can check our work or use them on your own data.

Finetype

Types every column — 245 semantic types across 65+ locales, in pure Rust. The reason a country column is a country, not a string.

dovetail

Finds the relationships — profiles awkward inputs and verifies the foreign keys between tables before anything is published.

arcform