超过6,000名开发者信赖

Turn Global News Into Actionable Business Signals

Track companies, competitors, people, brands and markets across global news. Filter on the entity itself rather than a keyword, narrow to the business events that matter, and push structured results into your own product, dashboard or AI agent.
免费试用-无需信用卡
额定5星 过了 848 用户
实时新闻文章177 国家
60 语言
  • CNN
  • Techcrunch
  • Vox
  • Apple
  • Microsoft
  • IBM
  • Bloomberg
  • Spotify

好处

Company events as structured data, not a reading list

A business signals API turns news coverage into structured evidence of what is happening to a company. APITube tags 8 million articles with 12 business event types — funding rounds, acquisitions, leadership changes, lawsuits and more — and resolves every company, person and brand to a single entity ID you filter on instead of a keyword.

1
12 business event types across 8M+ tagged articles
2
Filter by entity.id — one company, not every article containing its name
3
Sentiment scored per entity, not only per article
4
Boolean query language: combine events, entities and phrasing in one request
5
Restrict to authoritative publishers with source.rank.opr.min
6
300,000+ sources across 177+ countries and 59 languages
7
Delivery over REST, SSE, WebSocket, signed webhooks and a hosted MCP server
8
Export to CSV, XLSX, JSON, JSONL, XML, RSS or Parquet

Keywords find mentions. Entities reveal what actually matters.

Keyword alerts produce ambiguous matches and repeated coverage of the same story. APITube identifies the people, organizations, brands and locations inside each article, so you can monitor Apple Inc. rather than every article containing the word apple.

Keyword alerts

  • One word, many meanings — Apple the company, the fruit, the product
  • Every rewrite of the same announcement arrives as a new alert
  • A human has to open each article to judge relevance
  • No category, industry or sentiment attached to the match
  • An email digest is hard to build a product on

Entity-based monitoring

  • One numeric entity ID resolved to Wikipedia and Wikidata
  • Event type, industry, publisher rank and country narrow the set before you read it
  • Sentiment scored for the article and for the entity inside it
  • Related coverage for any article via /v1/news/story/{id}
  • JSON, streams and webhooks that feed an application directly
免费试用-无需信用卡
额定5星 过了 848 用户

发现可能性

我们的解决方案旨在帮助您实现目标。 无论你是小企业还是大公司,我们都有适合你的解决方案。

競爭性的

Competitive intelligence

Track launches, partnerships, funding and leadership moves across a named set of competitors.

Accounts

Company monitoring

Run continuous coverage on customers, suppliers, portfolio companies and strategic accounts.

Market

市场情报

Follow industries, technologies and regions through structured coverage instead of a clipping service.

投资

Investment research

Add company news, event context and sentiment to screening tools and research terminals.

Risk

Risk and compliance

Surface lawsuits, investigations, data breaches, recalls and bankruptcy coverage as they are reported.

代理人

AI agents

Give an agent entity-resolved news over MCP instead of unverified search snippets.

销售

Sales intelligence

Spot funding rounds, executive changes and expansion coverage that open a conversation.

通信

PR and reputation

Watch negative sentiment and fast-growing stories around a company, executive or brand.

Filters

Define exactly what counts as a signal

Combine entity search with structured metadata to cut irrelevant coverage and write a monitoring rule per company, market or use case.

Business event types

  • Lawsuits and investigations — 1.5M+ articles
  • Product launches — 1.4M+
  • Mergers and acquisitions — 1.1M+
  • Leadership changes — 1M+
  • Partnerships — 871K+
  • Funding rounds — 496K+
  • Recalls — 437K+
  • Earnings coverage — 396K+
  • Layoffs — 392K+
  • Data breaches — 112K+
  • IPOs — 74K+
  • Bankruptcies — 67K+

Entity filters

  • entity.id — the precise company, person or brand
  • organization.name, person.name, brand.name, location.name
  • Entity lookup via /v1/suggest/entities
  • 9 entity types, linked to Wikipedia and Wikidata
  • Mention frequency and position in the text
  • entity.sentiment.polarity — tone toward that entity
  • entity.sentiment.score.min and .max
  • ignore.entity.id to exclude a company
  • Entity blocks inside the boolean query language
  • Trending entities via /v1/news/trends

Context filters

  • industry.id — 1,264 industries
  • category.id — IPTC media topics
  • topic.id — 69 topics
  • language.code and source.country.code
  • source.domain and ignore.source.domain
  • source.rank.opr.min — Open PageRank 0 to 10
  • source.bias — left, center or right
  • sentiment.overall, sentiment.title, sentiment.body
  • published_at.start and published_at.end
  • author.name and title search with phrases

超过 300,000+ 资料来源

APITube受到世界各地团队的信任,帮助他们以前所未有的速度构建和交付令人惊叹的数字体验。

Last updated
6s ago
来源总数
512.645k
昨天的请求
27.062
昨天添加的文章
114.543k
文章总数
3.85b

常见问题解答

A business signals API returns news coverage of company events as structured data you can query and automate against. APITube classifies each article into an event type, resolves the organizations, people and brands mentioned in it to entity IDs, attaches sentiment, industry and publisher metadata, and returns the result as JSON over REST, a stream or a signed webhook.
An alert tells you a string appeared somewhere. A signal is a filtered, structured record: which company, which kind of event, which publisher, which country, which sentiment. That difference is what makes it programmable — this is the Google Alerts alternative for developers, meant to be consumed by an application rather than read in an inbox.
Entity-based monitoring filters on the real-world thing an article is about instead of on the characters in its text. APITube extracts entities from every article and links them to Wikipedia and Wikidata, so Apple Inc. carries a stable numeric ID. Filtering by that ID removes the namesakes, the fruit and the unrelated product line that a keyword query would return.
Yes. Resolve each competitor to an entity ID with /v1/suggest/entities, then query /v1/news/everything with entity.id plus the event types you care about — product-launch, partnership, funding-round, executive-change. Add source.rank.opr.min to keep the feed to serious publishers, and language.code or source.country.code to scope a market.
Nine types: person, location, organization, brand, product, event, disease, natural disaster and sport. Any of them can be filtered by numeric entity ID. Eight of the nine also accept a name filter — organization.name, person.name, brand.name, location.name, disease.name, disaster.name, event.name and sport.name — which must match the canonical name in the entity reference exactly.
Yes. language.code filters the language the article is written in, and source.country.code filters by the country of the publisher. Coverage spans 300,000+ sources in 177+ countries and 59 languages, so a restructuring reported only in German or a recall covered only in Japanese is reachable through the same filters.
Articles stay available through the API for 365 days from publication. One limit worth designing around: any request that searches text — the title parameter or the boolean query parameter — is capped to a 31-day window. Longer ranges are retrieved by filtering on entity, event type, industry, category or source rather than on words.
Yes. A hosted MCP server exposes news search and entity lookup as tools, so an agent can resolve a company to an entity ID and pull current, source-attributed coverage instead of guessing from search snippets. For direct integration there are SDKs for Node.js, Python, PHP and Dart, plus SSE and WebSocket streams.
It classifies, it does not conclude. Every article is scored against an event dictionary and tagged with its single highest-scoring event type — 44 types in total, 18 of them business events. There is no endpoint that hands back a finished signal object with the parties, the amount and the date extracted. You define what counts as a signal by combining event type, entity, sentiment and source filters, and your own logic decides what to do with the match.
It is a broad signal, not a precise label. Filtering by event.type alone returns loosely related coverage as well — executive-change on its own can surface a used-car listing. For a clean feed, pair it with headline search: a boolean query such as ("steps down" OR "appointed CEO") AND event.type:executive-change. Every demo on this page uses that pattern.
Three calls. Look the company up with /v1/suggest/entities?prefix=acme to get its entity ID, query /v1/news/everything with that entity.id plus your event types and filters, then either poll, open an SSE stream, or create a webhook in the dashboard so matches are pushed to your endpoint signed with HMAC-SHA256.
Yes, and no credit card is required. The free plan allows 100 requests per day with up to 10 articles per request, and truncates the article body to 200 characters. Every filter on this page, along with entities, sentiment, categories and industries, is returned in full on free — the limits are on volume, not on capability.

Business signals API for tracking company events in global news.

APITube classifies news coverage into business event types and resolves the companies, people and brands inside each article to a stable entity ID. Funding rounds, acquisitions, leadership changes, lawsuits, data breaches, recalls, earnings, layoffs, IPOs, bankruptcies, partnerships and product launches are filterable in the same request as industry, category, country, language, publisher rank and sentiment.

Keyword alerts return every article containing a word. Entity filtering returns coverage about the specific organization you named, which is what makes company monitoring worth automating. Resolve a name to its entity ID through the suggest endpoint, then combine that ID with the event types that matter to your team.

Competitive intelligence, sales intelligence, risk and compliance, investor research and PR teams all read the same underlying feed and cut it differently. A sales team watches funding rounds and executive changes; a compliance team watches lawsuits, investigations and data breaches; a communications team watches sentiment toward one brand across 59 languages.

Results arrive as JSON over REST, an SSE or WebSocket stream, or a signed webhook, and export to CSV, XLSX, JSONL, XML, RSS or Parquet. A hosted MCP server exposes the same search and entity lookup as tools, so an AI agent can pull source-attributed company news instead of guessing from search snippets.

相关解决方案

资源

开发者

我们用曲奇饼

通过单击"接受",您同意在您的设备上存储cookie以进行功能和分析。