Reuters Headlines & Stock Ticker Feed

Scrapes Reuters headlines with publication timestamps, images, language, and — for financial stories — the stock tickers each article is tagged with. Filter by ticker or language. No login, no API key.

Useful for financial news monitoring, ticker-linked sentiment research, and multilingual media tracking.

What makes this different

Stock tickers, straight from Reuters' own tagging. Every financial article arrives with the tickers Reuters editorially attached to it (NVDA.O, .FTSE, .GDAXI) as a real array, not a comma-joined string you'd have to split. That turns "find today's Reuters coverage of NVIDIA" into a single input field instead of keyword-matching headlines and hoping.

Six languages from one feed. Reuters publishes the same wire in English, German, French, Italian, Spanish and Portuguese. Each row carries its language, so you can either monitor one market or compare how the same story is framed across them.

Language detected from the URL, not the feed's own field. The feed reports en as the language for every article regardless of what language it's actually in — a German article and an English one both claim en. The URL path prefix is the only reliable signal, so that's what this Actor uses. Trusting the feed's own field would tag the whole dataset as English.

has_tickers as a one-field filter for separating financial coverage from general news, without inspecting the ticker array on every row.

Why this works when reuters.com doesn't

Reuters article pages reject anonymous clients outright. This Actor doesn't read them — it reads Reuters' own outbound syndication feed, published on the Arc XP platform specifically for aggregators to consume, which returns headline, timestamp, image, caption and tickers inline. The feed is the data source, not a pointer to a page.

The practical consequence: you get headline-level data (title, timing, tickers, image), not article body text. url is there for attribution and human follow-up.

Input

Field	Type	Notes
maxItems	integer	Deduplicated across pages. 0 walks the full ~5,000-article archive.
tickers	array	Optional. Reuters notation — NVDA.O, .FTSE.
locales	array	Optional. en, de, fr, it, es, pt.
proxyConfiguration	object	Defaults to Residential.

Output

{
  "article_id": "reuters.com:gold-slips-on-profit-taking-2026-08-25",
  "title": "Gold slips on profit-taking after rally to over three-month high",
  "published_at": "2026-08-25T11:41:23Z",
  "locale": "en",
  "stock_tickers": ["NVDA.O", ".FTSE"],
  "has_tickers": true,
  "image_url": "https://www.reuters.com/resizer/v2/...",
  "url": "https://www.reuters.com/markets/gold-slips-..."
}

Filters are applied after each page is fetched, so a narrow ticker filter reads more pages to fill your requested count. Each run also writes a RUN_COVERAGE record to the key-value store with the filters applied and how many articles came back.

Local development

pip install -r requirements.txt
python test_local.py --max 120 --out sample_output.json
python test_local.py --max 25 --locale en
python test_local.py --max 10 --tickers NVDA.O

sample_output.json is real output from a live 120-article run across six languages.