APITube turns the world's news into structured JSON. It aggregates articles from 500,000+ verified sources across 180+ countries in 60+ languages, and enriches every article before you see it: a sentiment score, extracted entities (people, organizations, locations, brands, events), IPTC categories, topics and industries. You call one HTTP endpoint and get data you can store, not HTML you have to parse.
One request instead of a crawler
Search lives at GET /v1/news/everything with 65+ filter parameters that combine in a single call: keyword in the title, language, country, source domain, source quality rank, sentiment polarity and score range, entity ID, media presence, published date range, sorting. Positive English coverage of OpenAI from the last seven days is one URL with an api_key on it. Filtering happens server-side, so you are not downloading a firehose and throwing 99% of it away.
Each article comes back with the fields you would otherwise reconstruct by hand: title and body, published and modified timestamps, canonical URL, source domain with a quality rank, language and country, images and videos, and per-entity sentiment alongside the article-level score. Entity-level sentiment means you can ask for stories where Tesla specifically is covered negatively, not only stories that happen to be negative overall.
Endpoints beyond search
/v1/news/top-headlines — breaking coverage by country, language or category
/v1/news/story — articles clustered into one story, so you count events instead of duplicates
/v1/news/trends — trending entities, topics and categories over time buckets
/v1/news/stream — a live Server-Sent Events feed of new matches
/v1/suggest/entities — turns a name like "Tesla" into the entity ID you filter on
Authentication is a single API key, passed as an api_key query parameter, an X-API-Key header or an Authorization: Bearer header. JSON is the default response; XML, CSV, TSV, JSONL, XLSX and Parquet exports come with paid plans. Webhooks and WebSocket dispatches push new matches to your endpoint as they are indexed, so a side project does not have to poll on a cron.
For AI side projects
APITube runs a hosted MCP server at https://mcp.apitube.io/ — nothing to install, no local process. It exposes search_news and suggest as tools, ships four prompts (monitor_company, topic_sentiment, breaking_news, compare_coverage), and works with Claude, Cursor, VS Code and other MCP clients. The MCP server is available on paid plans, starting with Starter.
Who builds on it
Brand and media monitoring: mentions across languages with sentiment already attached
Finance and market intelligence: news filtered by entity, industry and sentiment as a model input
RAG pipelines and AI agents: a live news corpus instead of scraped pages
Research and data science: historical coverage for backtesting and longitudinal work
Products that ship a news feed: apps, dashboards and newsletters without a crawler to maintain
Pricing
Free tier: 100 requests per day, 10 results per request, no card required. Starter is $29/month for 10,000 requests, Basic $99/month for 50,000, Professional $199/month for 150,000, with 20% off on annual billing. Pay-as-you-go costs $0.01 per request. SDKs for Node.js, PHP and Python live at github.com/apitube, and the documentation is at https://docs.apitube.io.
