Use Webz.io Contextual News Search as an Apify Actor. The Actor is webzio/news-search. It calls the News Search API (POST /api/news/context) once per run and writes one dataset row per article.
This is not a Python or npm package. You run it from Apify Console, from the Apify API, or as a tool on the Apify MCP server.
You pay on Apify. You do not need a Webz API token. The search is the same News Search API as the REST product.
A successful run charges:
search event, for the API call.apify-default-dataset-item event per article written to the dataset.Prices are on the Actor page. An invalid query is not charged.
recent developments on EU AI regulation.score, title, url, published_at, text, and source metadata such as domain, language, country, sentiment, and ticker.OUTPUT record for query and total_results.Pass Published within days or Published from, not both. Coverage is the last 30 days.
from apify_client import ApifyClient
client = ApifyClient("<APIFY_TOKEN>")
run = client.actor("webzio/news-search").call(run_input={
"query": "recent developments on EU AI regulation",
"k": 10,
"days": 7,
"language": ["english"],
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
print(item["score"], item["title"], item["url"])
pip install apify-client
Actor page: apify.com/webzio/news-search
Apify's MCP server exposes the Actor. Point an MCP client at:
https://mcp.apify.com?tools=webzio/news-search
The Actor uses Apify's default limited permissions, so it stays visible to that MCP server.
List filters take arrays of strings. Scalar filters take one value. Leave a field empty to skip it.
| Field | What it filters |
|---|---|
language | Article language, for example english |
country | Source country, ISO-2, for example US |
category | Article category |
sentiment | positive, negative, or neutral |
domain | Only these publisher domains |
exclude_domain | Skip these publisher domains |
topic | Topic tags |
person | People mentioned |
organization | Organizations mentioned |
location | Locations mentioned |
ticker | Stock tickers, for example NVDA |
political_bias | left, center, or right |
trust_category | trusted_news, fake_news, or satirical_news |
source_type | local_news, newsroom, or gov_news |
published_from / published_to | Publish dates, YYYY-MM-DD |
domain_rank_gte / domain_rank_lte | Tranco domain rank (1 is the most popular) |
score_gte / score_lte | Match score, 0–10. The API default floor is 4. |
allow_multiple_chunks_per_article | More than one excerpt from the same article |
Full filter reference: News Search filters.
k must be from 1 to 50.search event.ticker column: the API omits metadata that does not apply to that article. The field is still in the dataset schema.