Fintech
Machine-readable news for finance teams.
AI agents and RAG
A news tool with filters your model can call and JSON it can read, each article with a source and link.
Fresh, cited context for assistants, agents and RAG.
Teams putting a model in front of questions about the world.
Compact fields keep tool output short and cheap in context.
Objectives
0/3Answer current-events questions with citations.
Page deep through results with cursor pagination.
Retrieve over recent and historical news.
Summarise a topic's latest coverage on a schedule.
One article
CNBCA filtered search the model calls when it needs recent news.
Illustrative scenario
The model gets a news search tool. When a question needs recent information, it calls the tool, reads titles and descriptions, and cites a source and link for each claim.
At a glance
curl -G https://api.newsmesh.co/v1/search \ -d apiKey=$NEWSMESH_API_KEY \ --data-urlencode 'q=space telescope' \ -d category=science \ -d sortBy=relevant \ -d limit=5
import os, requestsr = requests.get( "https://api.newsmesh.co/v1/search", params={ "apiKey": os.environ["NEWSMESH_API_KEY"], "q": "space telescope", "category": "science", "sortBy": "relevant", "limit": "5", },)articles = r.json()["data"]
const params = new URLSearchParams({ apiKey: process.env.NEWSMESH_API_KEY, q: "space telescope", category: "science", sortBy: "relevant", limit: "5",});const res = await fetch(`https://api.newsmesh.co/v1/search?${params}`);const { data } = await res.json();
{
"data": [{
"article_id": "9e939fcbd7ed090236e52ad2473dcca1",
"title": "NASA's newest space telescope launches on a quest to explor…",
"source": "CNBC",
"category": "science",
"topics": ["space exploration", "space telescopes", "NASA"],
"people": ["Nancy Grace Roman"],
"country_relevance": ["United States"],
"link": "https://www.cnbc.com/2026/08/30/nasas-newest-telesc…"
}],
"next_cursor": "eyJwIjoi…"
}