Beacon
+ software+ startup− hiking
Name your company or product
Your company, your product, a competitor or the market you're in. When the name is a common word, add words a reporter would write next to it.
Where we listen
Where it lands
Developer tools
Agents
Every quarter hour, Rumoro checks the newest online news for your company, your product and your competitors. Each article that names you is scored, and the coverage worth acting on goes to Slack, a webhook or your AI agent, with the outlet and the reporter.
$5.80 on us · no card · no subscription
Each new batch of articles is read once and checked against all your keywords together.
A company quoted in the story matches too, when GDELT lists it among the article's organisations.
The reporter's name comes with the article when GDELT has it, so you know who to thank or pitch.
The links an article carries are kept, so you can see which coverage sends readers to you.
How it works
Add a keyword in the dashboard or with one API call. Articles published from then on start coming to you within the quarter hour.
Beacon
+ software+ startup− hiking
Your company, your product, a competitor or the market you're in. When the name is a common word, add words a reporter would write next to it.
GDELT GKG · 15 min → headline · names ∋ "Beacon"Each new batch of articles GDELT indexes worldwide, checked against every keyword at once and scored against how you describe your company.
Each alert sets its own relevance bar and filters on sentiment and intent, such as praise, complaints, comparisons or buying intent, then posts to Slack, Telegram, email, a webhook or your agent.
For founders and lean teams
Most media monitoring suites are built and priced for PR departments. When you just need to know which articles name you or a competitor, a keyword covers it, next to the forums and social posts your customers write.
Rumoro now lets AI agents read what customers say across 16 sources
northloop.press · share it, thank Hana OkaforThe same launch, rewritten from the press release
A syndicated copy · no actionSix listening tools for small teams, Rumoro among them
stackwire.news · check how you're describedPaperline answers with a price cut
A competitor · tell salesNews alerts
What to follow, what Rumoro checks and the alert your team gets. Pick one.
Search or monitor
Google Alerts emails you pages that match. GDELT turns the world’s online news into open data that anyone may use with a citation. Here is what each leaves to you. News data in Rumoro comes from The GDELT Project.
Coming from Google Alerts? See the full comparison →
What you get
Enough to know what ran, where and who wrote it before you open the article, and plainly what isn't read.
The headline, the link, when GDELT saw it, the sharing image and the outlet that published it.
The people and organisations GDELT found in the article, up to 30, which is how a company quoted in the story still matches.
The reporter when GDELT lists one, and up to 20 links from the article, so you can thank the writer and see who sends readers to you.
The article's text beyond those names, articles in other languages, and coverage from before you added the keyword.
Beyond alerts
See how much coverage you get against competitors, keep track of the outlets and reporters, and keep every keyword worth its cost.
Track competitors next to yourself and see who the press writes about more, week by week.
Every publication that writes about you is kept with your tags and notes, and every article it ran.
Bylines come with the articles, so a thank-you or a pitch goes to the person who wrote about you.
Narrow the feed to your domain and find the articles sending readers to your site, including the ones nobody told you about.
See which keywords bring real coverage and which mostly bring namesakes, with suggestions on what to exclude.
An hourly check for a sudden burst of articles, a run of negative coverage, noisy keywords or a channel that stopped delivering.
One daily or weekly round-up in Slack, Telegram or email, for teams that read coverage in one sitting.
Filter by sentiment, intent or outlet, save the view for your team and download a coverage report as CSV.
MCP, API, SDKs and CLI
Rumoro's MCP server gives Claude, Cursor or Codex the articles that name you, scored, with the outlet and the reporter. Your agent writes the coverage summary and drafts the thank-you notes for you to send.
News articles that link to your site, with a relevance of at least 60.
curl --get 'https://api.rumoro.dev/v1/mentions' \
--header "Authorization: Bearer $RUMORO_API_KEY" \
--data-urlencode 'platform=news' \
--data-urlencode 'minRelevance=60' \
--data-urlencode 'linkHosts=yourapp.dev'200Response · REST, SDKs and CLI
{
"data": [
{
"id": "a41f…",
"keyword": { "term": "rumoro" },
"post": {
"platform": "news",
"url": "https://northloop.press/2026/10/rumoro-ai-agents-customer-conversations-…",
"title": "Rumoro now lets AI agents read what customers say across 16 sources",
"text": "Rumoro now lets AI agents read what customers say across 16 sources\n\nNamed in the article: Rumoro, Paperline\n\nBy Hana Okafor · northloop.press",
"links": ["https://rumoro.dev/"]
},
"author": { "name": "northloop.press", "url": "https://northloop.press" },
"classification": {
"intents": ["praise"],
"relevance": 91,
"sentiment": "positive",
"note": "Launch coverage that reads well. Post it on your channels today and write the reporter a short note of thanks."
}
}
],
"nextCursor": null
}Coverage, alerts and cost
Create a keyword with your company's name. Rumoro checks the world's newly indexed English-language articles every 15 minutes, gives each match a relevance score, a sentiment and an intent, and routes the coverage you care about through your alert rules to Slack, Telegram, email, a webhook or your agent.
Rumoro reads the news through the GDELT Project, an open index of online news from around the world that publishes a new batch of articles every 15 minutes. Rumoro uses the English-language batches and cites GDELT on every news mention.
Partly. It matches the headline, plus every person and organisation GDELT picked out of the full text. A company quoted halfway down still matches when GDELT lists it, but a word that only appears in the running text does not.
Typically within a quarter hour of appearing in GDELT. GDELT releases one batch every quarter hour and Rumoro reads each batch once for all your keywords, so polling faster would not bring articles any sooner.
No. News is matched from the moment a keyword starts. Sources Rumoro can search, such as Reddit, GitHub and Stack Overflow, also bring a short history when you add a keyword, but news does not.
When GDELT has the byline, yes. The reporter's name appears with the outlet on the mention, which is the case for roughly a third of articles. The outlet itself is always shown as the author.
Yes, when GDELT lists the article's links, which it does for about four in ten articles. Rumoro keeps up to 20 per article, and you can filter your mentions by your domain.
For online news and the conversations around it, yes. Rumoro follows English-language web news next to Reddit, X, Hacker News and the other sources, scores every match and sends it where your team works. It does not cover print, TV or radio, and it is priced per keyword rather than per seat.
The GDELT Project is an open database that monitors news media worldwide and publishes what it finds as free data, updated every 15 minutes. Its terms allow commercial use with a citation and a link, which Rumoro includes on every news mention.
Many teams do. Google Alerts emails pages that match your words, as they appear or as a daily or weekly digest. Rumoro scores each article against your company, filters by intent and sends it to Slack, a webhook or your agent. The Google Alerts alternative page compares the two in detail.
Not yet. Rumoro reads GDELT's English-language batches, so articles published in other languages aren't matched.
Yes. Mute an outlet by its address, such as example.com, and its articles stay out of your feed and alerts from then on.
Yes. Connect Slack once and each alert rule posts matching articles to a channel with the headline, the outlet, the AI relevance score, a note on why it matters and buttons to open it, mark it handled or mute the outlet.
$5 per keyword a month, charged by the day, plus $0.008 per matched article, with news included next to every other platform the keyword listens on. There's no subscription to cancel, and the first $5.80 is on us.
Learn more
All sources
Pick the sources for each keyword. Matches from all of them share one score, one feed and one set of alerts, API and MCP tools.
Your first keyword is on us. $5.80 of credit, no card.