Skip to content

About

About ko.io

ko.io turns U.S. regulatory filings and federal data releases into clean, structured records, and serves them to people through a web terminal and to software through a REST API and an MCP server for AI agents.
Source-traced rows
100M+
Datasets
23
13F institutions
12,900+
REST endpoints
44
MCP tools
24
Free tier
200 calls / day

Why it exists

SEC EDGAR is the public record of institutional positions, insider trades and corporate events in the United States. It is also raw XML and SGML spread across dozens of form types: amendments replace originals without clear versioning, CUSIPs have no ticker mapping, and XBRL financials use different tags across companies and years.

Each of those problems is solvable; solving all of them, and keeping them solved every filing day, is an infrastructure project. ko.io does that work once so researchers, developers and their agents can query the result directly.

What is covered

Form 13F institutional holdings, Forms 3, 4, 5 and 144 insider filings, congressional trades disclosed under the STOCK Act, 13D/13G stakes, 8-K events, 10-K/10-Q financials, N-PORT fund holdings, fails-to-deliver and Reg SHO threshold lists, plus Treasury, Federal Reserve, BLS, CFTC, FDIC, EIA and OFR series. Each dataset's source, history, cadence and row count is listed on the dataset reference.

How the data is built

Scheduled pipelines download each source, parse it, validate it and load it into a columnar database. Data-quality tests (uniqueness, not-null, range, reconciliation) run on every build. Share classes are consolidated and managers that file under several CIKs are merged into one continuous portfolio.

Every SEC row keeps the accession number of the filing it came from. Nothing is estimated, interpolated or editorialized: no sentiment scores, no model-generated opinions.

  1. Source
    EDGAR filing
    The original SEC document
  2. Identifier
    Form, period, filer
    Which filing a row came from
  3. Structure
    Parsed row
    Normalised, linked fields
  4. Delivery
    Web · REST · MCP
    Same row everywhere
Every row names the form, period and filer it was parsed from.

Infrastructure

The database runs as a three-replica cluster. The API is served from Stockholm and Singapore, and a Cloudflare Worker routes each request to the nearer region. The terminal, the REST API and the MCP server read the same tables. Current health and 90-day uptime are on the status page.

Ways to use it

  • Terminal: browse institutions, stocks, insiders and members of Congress in the browser. No account needed.
  • REST API: 44 endpoints returning JSON, authenticated with an API key.
  • MCP server: 24 tools for Claude, Cursor, Windsurf, Zed, Codex and other MCP clients.
  • Playground: try API calls from the browser and export the results.

The free tier includes 200 API calls a day. Paid plans and their limits are on the pricing page.

Contact

Questions, partnerships or enterprise needs: admin@ko.io.

Privacy policyTermsGDPR

One dataset, three ways in

Start with the free tier

The free tier includes 200 API calls a day.