Skip to content

About ko.io

ko.io is the clean record of who owns what in U.S. markets: SEC disclosures parsed into structured data; 13F holdings reconciled against filing totals, with detected unit errors corrected. 5.9M+ SEC filings parsed, insider trades since 2006 and congressional trades since 2014, served to people, code and AI agents from the same tables.
SEC filings parsed
5.9M+
13F filings unit-corrected
11,587
Quarters of history
53
Datasets in the catalog
23
MCP tools
24
Free plan
200 calls / day

SEC filings parsed: distinct accession numbers, excluding XBRL-only references. 13F filings unit-corrected: selected filings whose dollars-versus-thousands unit the detector overrode, originals kept. Measured 2026-09-25.

Purpose

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 across dozens of form types, and it contains errors. In 13F alone, 11,587 of 311,477 selected filings reported their values in the wrong unit, a 1,000x difference; ko.io corrects them and keeps the originals. A reporting period can have an original filing and later amendments, CUSIPs need mapping to tickers, and a manager can file under several entities. ko.io does that cleanup once and checks each 13F institution-quarter against its own cover totals: 300,209 of 305,479 are published, the rest are quarantined with a reason. Researchers, developers and their agents query the result instead of repeating the work.

Sources and normalization

Form 13F holdings, Forms 3, 4, 5 and 144, congressional trades disclosed under the STOCK Act, 13D / 13G, 8-K buybacks, 10-K / 10-Q financials, N-PORT, fails-to-deliver and Reg SHO lists, plus Treasury, Federal Reserve, BLS, CFTC, FDIC, EIA and OFR series. Each dataset's source, history, cadence and access is in the dataset catalog.

Pipelines download, parse, test and load each source; share classes are consolidated and a manager's linked filing entities are combined into one portfolio where the mapping exists. SEC datasets are parsed from public EDGAR filings. 13F, Form 4 and similar rows don't carry accession numbers; the filings endpoints return each filer's source documents. Reported values and calculated measures are labeled separately. The full method is under how the data is built.

The Terminal, the REST API and the MCP server read the same tables, served from data centers in Europe and Asia-Pacific over a global edge network spanning 340+ cities, with precomputed research views for large historical queries. Current health is on the status page.

Ways to use the data

  • Terminal: browse institutions, stocks, insiders and members of Congress in the browser. No account needed.
  • REST API: 44 core endpoints returning JSON (full list in the API reference), authenticated with an API key.
  • MCP server: 24 tools for Claude Code, Codex, Cursor, Grok (xAI API), Gemini CLI and other MCP clients.

The Free plan includes 200 API calls a day for the most recent quarter; full history and the other plans are on the pricing page.

Screens of ko.io captured 2026-09-24.

Contact

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

ko.io is operated by ko.io, a company organized under the laws of the European Union, as set out in the Terms of Service.

Privacy policyTermsGDPR

Start free

Free plan: 200 calls a day over REST and MCP (one shared quota), 500 rows per request, most recent quarter.