financial

SEC EDGAR Scraper — Filings, Full-Text & XBRL Financials

Extract SEC EDGAR filing metadata, full-text 10-K/10-Q sections, EDGAR full-text search results, and XBRL financial facts as clean, RAG-ready JSON. No API key required.

TypeScript Cheerio United States

Try on Apify

Key Features

Four modes in one actor — filing metadata, full document text, EDGAR full-text search, and XBRL financial facts

RAG-ready chunking — section, paragraph (~2000 chars), or none; every chunk tagged with its source Item and order

Automatic 'Item N' section parsing for 10-K/10-Q (Item 1A Risk Factors, Item 7 MD&A, and more)

XBRL facts with taxonomy, tag, label, unit, value, fiscal year/period, form, and accession number

Fact deduplication collapses XBRL restatements to one row per period, keeping the earliest disclosure

Company resolution by ticker, CIK, or name against SEC's official ticker file, with fuzzy fallback

EDGAR full-text search with quoted phrases, form-type, and date-range filters (2001 → present)

SEC-compliant rate limiting and User-Agent — no API key and no login required

Use Cases

Financial RAG / LLM pipelines needing section-chunked, embedding-ready 10-K and 10-Q text
Investment research — pull revenue, net income, and diluted EPS time series as clean XBRL rows
Compliance & ownership monitoring — Form 4, SC 13D/13G, and risk-factor language across companies
Fintech products that need structured EDGAR data without building a custom crawler
Market and industry research via full-text search — who discloses a phrase, and since when
BI and spreadsheet workflows consuming clean XBRL financial fact rows

Input Parameters

Parameter	Type	Required	Description
`mode`	string	Yes	Extraction mode: filings, fulltext, search, or facts (default: filings).
`ticker`	string	No	Stock ticker, e.g. AAPL. One of ticker/cik/companyName is required at runtime.
`cik`	string	No	SEC Central Index Key, e.g. 320193 — takes precedence over ticker and companyName.
`companyName`	string	No	SEC registrant name, exact or fuzzy-matched against the official ticker file.
`query`	string	No	EDGAR full-text search expression, e.g. "supply chain disruption" — required for search mode.
`formTypes`	array	No	Form types to include, e.g. 10-K, 10-Q, 8-K, S-1, DEF 14A, Form 4. Empty = all forms.
`chunking`	string	No	Fulltext RAG strategy: section, paragraph (~2000 chars), or none (default: section).
`maxItems`	number	No	Max items saved; each saved item is one billed event (default: 100).

Output Example

 1{
 2  "itemType": "filing-fulltext",
 3  "title": "Apple Inc. — 10-K 2025-10-31",
 4  "company": "Apple Inc.",
 5  "ticker": "AAPL",
 6  "cik": "0000320193",
 7  "formType": "10-K",
 8  "filedAt": "2025-10-31",
 9  "accessionNo": "0000320193-25-000123",
10  "documentUrl": "https://www.sec.gov/Archives/edgar/data/320193/...",
11  "sections": [
12    { "name": "Item 1A — Risk Factors", "charCount": 38241 },
13    { "name": "Item 7 — Management's Discussion and Analysis", "charCount": 21077 }
14  ],
15  "chunks": [
16    { "text": "The Company's business, reputation, results of operations...", "section": "Item 1A — Risk Factors", "order": 12 }
17  ],
18  "textLength": 220151
19}

Pricing

Pay-per-event — you’re billed only for items actually saved:

Event	Price	What it covers
Filing metadata / search hit	$0.001	One filing metadata record or full-text search result
Full-text filing	$0.005	One filing extracted, section-parsed, and chunked for RAG
XBRL fact	$0.0002	One XBRL financial fact row

A 1,000-filing metadata pull is ~$1.00; 1,000 XBRL facts is ~$0.20. maxItems caps both volume and cost.

Tips

Start with facts mode for financials. Pulling XBRL rows (revenue, net income, EPS) is far cheaper and cleaner than parsing full filings when you only need the numbers.
Use chunking: section for RAG. It keeps each Item (Risk Factors, MD&A) intact so retrieval returns coherent, citable passages.
Full-text search covers 2001 onward. For older disclosures, resolve the company by CIK and pull filing metadata directly.

Frequently Asked Questions

Do I need an SEC API key?

No. SEC EDGAR is free public data. The actor identifies itself with a compliant User-Agent and self-throttles under SEC's rate limit automatically — no key and no login.

How far back does the data go?

EDGAR full-text search (EFTS) covers filings from 2001-05-04 onward and is capped at 10,000 hits per query. Filing metadata goes back to 1994, and XBRL financial facts cover everything the company has reported in XBRL.

Is the output ready for LLMs and RAG?

Yes. In fulltext mode the actor parses 10-K/10-Q into 'Item N' sections and emits embedding-ready chunks (section or ~2000-char paragraph), each tagged with its source Item and order index.

Why are some filings skipped in fulltext mode?

Filings whose primary document isn't HTML or TXT (for example, XBRL-only Form 4 XML) can't be section-parsed, so they're skipped — and not charged.