📑 SEC Filings to Markdown for RAG · EDGAR → LLM-Ready

Convert SEC EDGAR filings into clean, chunked, citation-tagged Markdown — built for AI engineers feeding financial filings into RAG pipelines and LLM agents.

⚡ What you get

Field	Description
company / cik / ticker	Issuer identity
form	Filing type (10-K, 10-Q, 8-K, 13F, …)
filingDate	Date filed
accessionNumber	SEC accession number (citation)
sourceUrl	Direct link to the source document (citation)
chunkIndex / totalChunks	Position within the filing
markdown	Clean Markdown chunk, ready for embedding

🎯 Use cases

1. AI engineers building financial-research copilots / RAG over filings
2. Quant & fundamental analysts loading 10-Ks into a vector store
3. Compliance teams building searchable filing knowledge bases
4. Fintech products needing LLM-ready filing text with citations

🚀 Sample inputs

{ "companies": ["AAPL","MSFT"], "formTypes": ["10-K","10-Q"], "maxFilingsPerCompany": 2, "chunkWords": 800 }

{ "companies": ["NVDA"], "formTypes": ["8-K"], "maxFilingsPerCompany": 5 }

{ "companies": ["320193"], "formTypes": ["13F-HR"] }

📦 Sample output

{ "company": "Apple Inc.", "cik": "0000320193", "ticker": "AAPL", "form": "10-K",
  "filingDate": "2025-11-01", "accessionNumber": "0000320193-25-000123",
  "sourceUrl": "https://www.sec.gov/Archives/edgar/data/320193/.../aapl-20250927.htm",
  "chunkIndex": 0, "totalChunks": 42, "markdown": "# Item 1. Business\nThe Company designs..." }

📊 Sample Output

🛠 How it works

1. Source — resolves tickers→CIK and reads filings from the official SEC EDGAR APIs (data.sec.gov, www.sec.gov/Archives).
2. Parser — strips scripts/styles and converts filing HTML to ATX Markdown.
3. Chunking — splits into ~chunkWords-word chunks for embedding.
4. Schema — one row per chunk with full citation fields (accession + source URL).
5. Fallback — unresolved tickers / failed docs are logged and skipped; the run still succeeds.

🔗 Related Actors

• SEC EDGAR Scraper — structured filing data
• SEC Form 13F Holdings Tracker — institutional holdings
• RAG Web Browser — web content for retrieval
• Website Content Crawler — full-site Markdown for AI

💰 Pricing Example

Pay-per-event: $0.005 per run + $0.04 per Markdown chunk (document-record).

Chunks	Cost
100	~$4.00
500	~$20.00
2,000	~$80.00
Apify's $5 free credit covers ~124 chunks. Start free →

⚖️ Legal & data sources

Data is from the public SEC EDGAR system (data.sec.gov, www.sec.gov) — U.S. government public-domain filings. Requests use an identified, contact-bearing User-Agent per SEC access guidance. You are responsible for your downstream use.

❓ FAQ

Which forms are supported? Any EDGAR form type — pass them in formTypes (10-K, 10-Q, 8-K, 13F-HR, DEF 14A, …). Ticker or CIK? Either; tickers are resolved to CIK automatically. Are citations included? Yes — every chunk carries the accession number and source URL. How big are chunks? ~chunkWords words (default 800); tune for your embedder. Is the data fresh? Pulled live from EDGAR at run time. Cost control? Use maxFilingsPerCompany and formTypes to bound output.

🆘 Troubleshooting

• Company not found — check the ticker symbol or pass the CIK directly.
• 0 chunks — the requested formTypes may not exist in that issuer's recent filings.
• Huge output — lower maxFilingsPerCompany or narrow formTypes.
• Markdown noise from exhibits — narrow to the primary form types you need.

🏷️ About NexGenData

NexGenData builds structured public-data tools for analysts, developers, and operators. Full catalog: thenextgennexus.com.