A news data API with typed, consistent fields
News as structured data you can query: numbers with units, people, organizations, places, category, importance and dates on every story. Here is every field.
The fields on every story
A news data API where the structure is the product. Every field below comes back on every story.
| Field | Type | Example |
|---|---|---|
id | integer | 13023 |
title | string | Global economic outlook projects lower oil prices ... |
language | string | en |
facts | string[] or null | S&P Global expects Brent crude to drop to 70 dollars ... |
category | string | finance |
topics | string[] | ["inflation","interest rates","energy"] |
people | string[] | ["Ken Wattret"] |
organizations | string[] | ["S&P Global","Federal Reserve"] |
places | string[] | ["United States","Eurozone","Japan"] |
countries | string[] | ["United States","Japan"] (countries of the places, or of the main organizations when no place is named) |
regions | string[] | ["Middle East"] (regions such as Europe or the Gulf; ["Global"] when the story names no place) |
numbers | object[] | {"value":70.0,"unit":"USD","what":"Brent crude price forecast ...","value_normalized":70.0,"unit_code":"USD"} |
event_date | string or null | 2026-11-19 |
importance | integer 0-100 | 48 |
outlet_count | integer | 1 |
source_languages | string[] | ["tr"] |
first_published, updated_at | ISO 8601 string | 2026-11-19T13:49:02Z |
Numbers as data
/v1/facts searches the numbers across all stories. Each hit carries the value, unit, a normalized value, what it measures and the story it came from. Filter with min and max.
curl -H "X-API-Key: YOUR_KEY" \ "https://api.latestnewsapi.com/v1/facts?what=inflation&unit=percent&min=4&limit=2"
import requests r = requests.get("https://api.latestnewsapi.com/v1/facts", headers={"X-API-Key": "YOUR_KEY"}, params={"what": "inflation", "unit": "percent", "min": 4}) for f in r.json()["data"]: print(f["value"], f["what"])
{
"data": [{
"value": 4.0, "unit": "percent",
"what": "annualized core inflation of the last three months in the US",
"value_normalized": 4.0, "unit_code": "percent",
"story_id": 13023,
"story_title": "Global economic outlook projects lower oil prices ...",
"category": "finance",
"first_published": "2026-11-19T13:49:02Z"
}, ...]
}Six data pulls to copy
Slim records for a table
fields=id,title,categoryfacts=noneStories with numbers only
has_numbers=truecategory=financeAutocomplete organizations with story counts
type=organizationqRank by importance for a feed
sort=-importancemin_importance=60Get one story by id
idFind related stories
relatedBuilt to be parsed
Typed and consistent
Same fields on every story, in every category. Empty lists stay lists, unknown dates are null.
Machine-readable docs
OpenAPI, Markdown docs and llms.txt are all public.
Credits that are easy to model
1 credit per read, 25 for an article, 0 for usage and categories. Failed calls cost 0. See pricing.
Turn any story into a readable article
On paid plans, one call writes an article from a story's facts in any of 45 languages. Facts and our own titles, never copied article text.
GET /v1/stories/13023/article?lang=de # 25 credits, en tr de fr es it pt nl pl sv ru uk ar fa hi id ja ko zh
News data API questions
What is a news data API?
An API that returns news as structured data instead of text to scrape. LatestNewsAPI gives titles, fact lines, numbers with units, people, organizations, places, category, importance and dates.
What types are the fields?
See the field table on this page. Lists such as people and places are arrays of strings, numbers is an array of objects, and importance and outlet_count are integers.
Can I search numbers directly?
Yes. /v1/facts searches numbers by topic, unit and range, for example what=inflation&unit=percent&min=40. Values come with value_normalized and unit_code.
Can I choose which fields come back?
Yes. Use fields= to list the ones you want and facts=none, top3 or all to control fact lines. Smaller responses are faster.
Is there an OpenAPI spec?
Yes, at /openapi.json. Docs are also available as Markdown at /docs.md and an llms.txt at /llms.txt.
What is event_date?
The day the thing happened (YYYY-MM-DD), or null if unknown. It differs from first_published, which is when we first saw the story.
Does it include original article text?
No. Facts and our own titles, never copied article text.
See the data in your own code
Free plan, 100 requests a day. The same fields come back on every plan.