Today, we're introducing news search across our index of over 200 million articles. You can search for a topic or find news about a specific company, including announcements, press releases, and reporting.
LLMs have a tendency to act like they know things when they don't. If you ask about a company, you often get a confident answer that was right maybe six months ago. The company may have since launched a product, raised a round, or been acquired. The model doesn't necessarily know exactly what it doesn't know.
Agents need a way to check for the latest information on the web. News search gives them the latest available data, so they can look up what's changed before they answer, double-check what they know, and research more effectively.
Find out what's happening
There are two ways to search:
- You can just describe what you're looking for in plain language, like "companies deploying AI agents for enterprise customer support."
- You can identify a company by its name, domain, ticker, or ISIN.
A company lookup can be useful when your agent already knows who it's researching. For example, it can find recent announcements for a customer call. When updating a company record, it can check for new products or acquisitions. For a briefing, it can pull together reporting on the topic and link back to the sources.
With Context, you can also narrow results by publisher, language, publisher country, or article type, as well as the publication date range. You can sort company news by newest or relevance, and topic searches are ranked by relevance.
Try it
Here's a request for the latest press releases about Stripe. Store your API key in CONTEXT_DEV_API_KEY and run:
curl https://api.context.dev/v1/news/search \
--request POST \
--header "Authorization: Bearer $CONTEXT_DEV_API_KEY" \
--header "Content-Type: application/json" \
--data '{
"searchBy": {
"type": "entity",
"entity": { "type": "domain", "domain": "stripe.com" }
},
"filterBy": { "articleType": ["press_release"] },
"sortBy": { "type": "newest" },
"limit": 10
}'For a topic search, use this request body instead:
{
"searchBy": {
"type": "natural_language",
"query": "companies deploying AI agents for enterprise customer support"
},
"limit": 10
}Search also pairs well with our Highlights API, which gives agents just the context they need for their task.
This is just another step towards giving agents the information they need to understand the world as it is now.
Try news search in the dashboard, read the API reference, or use it through the Context MCP server.