CNBC Scraper

Collect complete publicly available CNBC articles for news monitoring, market research and media analysis. Start with an article URL, a section page, the home page or an official RSS feed. Feeds and section pages discover articles; each saved row contains the public article body.

What you receive

Each article includes its headline, description, complete public text, authors and author links, categories, keywords, publication and update dates, image links and visible image context, canonical URL and word count when the source provides them. Publisher and public video metadata are included when present. Missing optional values are omitted or represented by empty arrays. Image links point to the original images.

The dataset contains articles only. The run summary reports saved counts, failed requests and source notices, with run times and source URLs in the key-value store. If an article is saved before a later request fails, the run keeps that article and reports partial completion. A run without any complete public articles reports failure. Subscriber pages and access challenges are not scraped.

Start a run

1. Add public CNBC URLs in the input tab.
2. Choose the maximum number of articles; the default is 10.
3. Optionally set a publication date such as 2026-10-01 or 7 days.
4. Enable Only new articles to skip URLs saved by earlier runs of this Actor.
5. Start the run, then download the dataset as JSON, CSV or Excel.

Example input:

{
  "startUrls": [
    {
      "url": "https://www.cnbc.com/2026/10/06/treasury-yields-fed-fomc-minutes.html"
    }
  ],
  "maxArticlesPerCrawl": 2
}

Verified output example

This record was retrieved from the public CNBC article on October 6, 2026. The content below shows a 14-word opening excerpt; the actual record contains the complete public text, and wordCount describes that complete text.

{
  "articleId": "world-bank-east-asia-growth-inflation-ai-exports-",
  "url": "https://www.cnbc.com/2026/10/06/world-bank-east-asia-growth-inflation-ai-exports-.html",
  "title": "World Bank raises East Asia growth outlook to 4.5%, flags AI risks",
  "content": "The World Bank has raised its growth forecast for the East Asia and Pacific",
  "authors": [
    "Anniek Bao"
  ],
  "category": "Economy",
  "publishedAt": "2026-10-06T04:06:57.000Z",
  "updatedAt": "2026-10-06T04:07:24.000Z",
  "wordCount": 602,
  "publisher": { "name": "CNBC", "url": "https://www.cnbc.com/" },
  "authorDetails": [{ "name": "Anniek Bao", "url": "https://www.cnbc.com/anniek-bao/" }],
  "videos": [{ "name": "Sam Altman: The world should accept some bad things happening for the benefits of AI", "duration": "PT3M56S" }]
}

When a detail-page sourceUrl would repeat the canonical url, it is omitted. A distinct listing-page source is retained.

Output update

url identifies the article. articleId is the article's URL filename slug (without .html), retained under the existing field name for compatibility; CNBC's page URL is its source. keywords is the sole field for the source's native keyword list, including its original order and repeated terms. The duplicate tags alias is removed; migrate consumers to keywords. The categories array keeps distinct native labels in source order, while category remains the first label for compatibility. Derived sourceDomain and collection details such as extractionMethod, transport and scrapedAt are no longer included in article rows. Run times, source URLs and timestamped source notices remain in the key-value run summaries.

Filters and limits

startUrls accepts up to 50 public CNBC URLs. urls, start_urls, url and maxItems remain available as compatibility aliases. The article limit can be set from 1 to 500. Date filtering uses the source's publication timestamp when available.

Requests have a 30-second timeout and a 4 MB response limit by default. You can adjust these in the input tab. Temporary network errors receive at most one retry; access challenges stop further source requests. Only-new history retains the latest 5,000 saved article URLs.

Cost and support

The current price appears in the Actor's pricing tab. Runtime and data transfer depend on the number and size of pages fetched. Focused article URLs and a small article limit help keep runs short.

For questions or problems, use the Actor's Issues tab and include the run URL.