US Congressional Record Scraper - Floor Speeches, Statements & Legislative Text

Extract daily floor speeches and statements from the US Congressional Record via the official Congress.gov API. Returns per-article records with section, volume, issue metadata, page numbers, and optional full plain text. Coverage spans from 1995 to the present across all CR sections: Daily Digest, Senate, House, and Extensions of Remarks.

What data does it extract?

Each record represents a single article from a daily Congressional Record issue:

Field	Description
congress	Congress number (e.g. 119)
session	Session number within the Congress
volume	CR volume number
issue_number	Issue number within the volume
issue_date	Publication date (YYYY-MM-DD)
section	Section name: Daily Digest, Senate Section, House Section, Extensions of Remarks Section
article_title	Title of the article or speech
start_page	First page in the printed Record
end_page	Last page in the printed Record
article_text	Full plain-text body (populated when includeFullText: true). Capped at 50,000 characters.
pdf_url	URL to the PDF version of this article
source_url	Canonical URL on congress.gov
scraped_at	ISO-8601 scrape timestamp

Text availability note: article_text is populated only for articles that have an associated Formatted Text URL in the API. For most pre-2000 records, only PDF versions exist — article_text will be null even with includeFullText: true. Full text coverage is reliable from approximately 2000 onward.

What does the scraper do?

It calls the Congress.gov API to list CR issues by date or Congress number, then fetches each article's detail record. With includeFullText: true, it makes one additional HTTP request per article to retrieve the plain text body. The default request delay (400 ms between requests) keeps throughput well within the 5,000 req/hr limit on a free API key.

How to use it

Field	Type	Default	Description
maxItems	integer	15	Maximum number of article records to return
dateFrom	string	—	Start date (YYYY-MM-DD)
dateTo	string	—	End date (YYYY-MM-DD)
congress	integer	—	Filter by Congress number (e.g. 119 for the 119th Congress, 2025-2027)
includeFullText	boolean	true	Fetch the full plain-text body of each article
apiKey	string	—	Your free api.congress.gov API key. If blank, a shared key is used (shared rate limit).

Get a free API key at api.congress.gov/sign-up/ — instant issuance, no review required. A production key allows 5,000 req/hr.

Fetch recent articles (metadata only)

{
  "dateFrom": "2026-05-15",
  "dateTo": "2026-05-15",
  "includeFullText": false,
  "maxItems": 100
}

Download full text from a specific Congress

{
  "congress": 119,
  "dateFrom": "2026-01-01",
  "dateTo": "2026-03-31",
  "includeFullText": true,
  "maxItems": 1000
}

Historical backfill

{
  "dateFrom": "1995-01-04",
  "dateTo": "1995-12-31",
  "includeFullText": true,
  "maxItems": 5000
}

Use cases

• Legislative NLP training corpus — collect floor speeches with section labels and date metadata to train or fine-tune models on legislative language
• Lobbying analytics — track which bills are discussed on the House or Senate floor, correlated with sponsor party and subject
• Committee hearing archive — build a searchable archive of Extensions of Remarks statements for specific policy areas
• Legislative monitoring — run daily with dateFrom set to the previous day to ingest each new CR issue as it publishes
• Journalism and data reporting — export historical floor speech text for topic modeling or legislator activity analysis

FAQ

Is the Congress.gov API free? Yes. A free key allows 5,000 req/hr. Without a key, a shared demo key is used, which is limited to 30 req/hr and suitable for testing only.

Why is article_text null for some records? Articles that exist only as PDFs in the Congress.gov system do not have a Formatted Text URL. This is most common for records before 2000. The PDF link is always included in pdf_url.

What is a typical run size? Each daily CR issue contains roughly 100 articles across all four sections. A full month of articles is approximately 2,000 records.

Output is available in JSON, CSV, and Excel via the Apify dataset export panel.