Worldwide News Search by Keyword

Search news coverage from tens of thousands of outlets in more than 65 languages, and get each article's headline, link, outlet, outlet country, language, date, image and tone score in one clean table. Data comes from the GDELT Project, an open, free database that monitors world news media around the clock and publishes it for public and automated use. No Google News scraping, no logins.

Who uses it

• PR and communications teams tracking mentions of a brand, executive or product across world media.
• Investors and analysts following news on companies, sectors, central banks or commodities.
• Researchers and journalists studying how a topic is covered by country, language or outlet.
• Risk and security teams monitoring news about places, suppliers or events.

How matching works

By default the tool scans GDELT's open 15-minute news files for your time window. An article matches when your terms appear in its headline, its URL, or the people, organizations, places and topic tags GDELT extracted from the full article text. Because GDELT translates foreign-language coverage, an English search such as "artificial intelligence" also finds German, Spanish or Hindi articles about the topic.

Optionally, the GDELT search API engine first tries GDELT's own full-text search. That service limits cloud servers heavily, so when it is unavailable the run continues automatically with the news files.

Input

Field	What it does
Search queries	One search per line. Quotes for exact phrases ("interest rates"), OR between alternatives (tesla OR rivian), parentheses for groups and -word to exclude.
Time window	6h, 24h, 3d, 2w and so on. Or set a start and end date (UTC).
Source countries	Outlet country, as a name without spaces (unitedstates, germany) or a two-letter FIPS code (US, UK, GM).
Source languages	english, spanish, french, arabic, chinese and more.
Only these domains / Exclude domains	Limit to or drop specific outlets, e.g. theguardian.com.
Maximum articles per query	Cap per search (default 100).
Sort	Newest first (default) or oldest first.
Remove duplicate headlines	Drop syndicated copies with the same headline. Duplicate URLs are always removed.
Fetch article description	Optional. Visits each article to add its meta description, first paragraphs, author and publish time.
Maximum days to scan	Upper limit on the window for the news-file engine (default 3, up to 31).

Output

query, title, url, domain, sourceCountry, language, seenDate (when GDELT first saw the article, UTC, 15-minute precision), image, tone (GDELT average tone, negative to positive, usually between -10 and +10), wordCount, persons, organizations, locations, and with article details on: description, leadText, author, publishedTime, detailsError.

Typical use

Schedule a run every hour with a 1h or 2h window and your brand names, then send new articles to Slack, email or Google Sheets with Apify integrations.

Notes and limits

• Newest-first searches stop as soon as enough articles are found, so they are fast. Long windows with rare terms read more files and take longer.
• Outlet country comes from GDELT's published outlet-to-country list; some aggregators are listed under the country where they are based.
• Common words also match GDELT's topic tags, which can include articles that only touch on the subject. Use exact phrases and exclusions to narrow results.
• Article details are fetched politely from public pages. Some sites refuse automated visits or are paywalled; those fields then stay empty and the article is still returned.
• Headlines and links belong to their publishers. GDELT data is free to use with attribution to the GDELT Project.

Plot Point Technologies is not affiliated with the GDELT Project or any news publisher.