# Data sources and attribution

Where every figure comes from, which sources are left out, and how to credit the data when you show it to others.

The API serves only data we are allowed to pass on: public records, openly licensed sources, and what we derive from them. Each answer says where it comes from, and some sources ask for a notice when their data is shown.

## Sources

| Source | What the API serves from it | Tools (examples) |
|---|---|---|
| **SEC EDGAR** | The company directory; the index of every filing; financial statements from XBRL filings, normalized onto one chart of lines and as filed; insider transactions (Forms 3, 4 and 5); institutional holdings (Form 13F); executive pay from proxy statements | `search_companies`, `get_company`, `get_financials`, `list_filings`, `get_insider_trades` |
| **U.S. House Clerk and Senate** | Members' periodic transaction reports | `get_congress_trades` |
| **congress-legislators** (public domain) | Members' party, state and chamber | `get_congress_trades` |
| **The GDELT Project** | Press headlines found in GDELT's news index: headline, outlet, date and link, never the article text | `get_headlines` |
| **FRED** (Federal Reserve Bank of St. Louis) | U.S. economic indicators, their history and release dates | `get_economy` |
| **FINRA** | Consolidated short interest by settlement date | `get_short_interest` |
| **Fundalyze** | What we derive from the sources above: standardized statements, margins and growth, founder and control research, peer lists, AI summaries of filings | several |

## What is not in the API

No live or delayed prices, no market capitalisations or valuation ratios, no analyst estimates, and no data from commercial market-data vendors, whatever the website shows to subscribers. Those sources do not allow us to pass their data on.

## Notices that travel with the data

Some answers carry a notice. When you show that data to someone else, show the notice with it.

- **FRED:** every `get_economy` answer ends with "This product uses the FRED® API but is not endorsed or certified by the Federal Reserve Bank of St. Louis." Some series belong to their publishers (Freddie Mac, the University of Michigan, Cboe and others); for those, the answer includes the citation to show (`citation` in `data`).
- **The GDELT Project:** headlines are credited to The GDELT Project (the `attribution` object in `data`). Each headline belongs to its publisher and links to the article on the publisher's site.
- **FINRA:** every short-interest answer carries this line in `text` and an `attribution` object in `data`: "Short interest data are owned by and sourced from FINRA (Financial Industry Regulatory Authority, Inc.), including figures derived from it; FINRA does not endorse Fundalyze."
- **SEC filings** are public records; filing links point to sec.gov.

Our [Data attributions](/attributions) page lists every source's terms.

## Crediting Fundalyze

When output from the API is shown to other people (in a post, a report, a screenshot, an answer your assistant writes for someone else), credit it as **"Data via Fundalyze"** with a link to fundalyze.ai or to the page the answer links to. Every answer includes that link: `More: <link>` at the end of `text`, and `link` in REST responses.

## What you may do with the data

The [API and MCP access](/terms#api) section of our Terms sets the rules. In short:

- **Yes:** use what the API returns for your own analysis and in your own tools, including AI assistants and agents you run for yourself, and keep what you need for that.
- **No:** resell, sublicense or redistribute the data, in raw or bulk form; offer it to others as a feed, a dataset or an API of your own; use the API to rebuild our dataset or a large part of it; share a key, or use more keys or accounts to get around the limits.

Your access ends when your plan does.