Real-time news
Articles as they publish, filtered the way you need.
Search
Search with phrases, exclusions and boolean operators, then filter by category, country, source and date.
Precise queries without building a search engine.
Anyone looking for specific coverage rather than the latest.
The operators work together in a single q parameter.
What you get
0/3Track brands, people and products.
Collect every article on a topic.
Give the model a search tool it can call.
Find earlier coverage before writing.
Query syntax
q parameterOne query, filtered to a category and ranked by relevance.
Example
A team searches technology news for AI stories, leaves out opinion pieces, and ranks the results by relevance.
At a glance
curl -G https://api.newsmesh.co/v1/search \ -d apiKey=$NEWSMESH_API_KEY \ --data-urlencode 'q=AI -opinion' \ -d category=technology \ -d sortBy=relevant \ -d limit=10
import os, requestsr = requests.get( "https://api.newsmesh.co/v1/search", params={ "apiKey": os.environ["NEWSMESH_API_KEY"], "q": "AI -opinion", "category": "technology", "sortBy": "relevant", "limit": "10", },)articles = r.json()["data"]
const params = new URLSearchParams({ apiKey: process.env.NEWSMESH_API_KEY, q: "AI -opinion", category: "technology", sortBy: "relevant", limit: "10",});const res = await fetch(`https://api.newsmesh.co/v1/search?${params}`);const { data } = await res.json();
{
"data": [{
"article_id": "5a99ebb20294d429f368f9a4e21bf83c",
"title": "Nvidia launches new platform for reining in rogue AI agents",
"source": "TechCrunch",
"category": "technology",
"topics": ["AI technology", "software development", "technology trends"],
"people": ["Jensen Huang"],
"country_relevance": ["United States"],
"link": "https://techcrunch.com/2026/09/28/nvidia-launches-n…"
}],
"next_cursor": "eyJwIjoi…"
}