Skip to main content
Easily find the most relevant information from trusted sources and your own data. Use it to power agents that give accurate, real-time answers. Ready to start using the Search Service? Let’s dive in!
We will create a sample API request and describe the main request parameters. The best way to try it out is the Developer Platform > Search PlaygroundThe query request object contains the following main parameters:
  • text: Natural-language search string. Not mandatory when filtering by entities, keywords or other filters. Mutually exclusive with texts.
  • texts: Array of up to 5 search strings for multi-text search. Use when a single query cannot capture all aspects of your intent. Mutually exclusive with text. When the reranker is enabled and more than one text is provided, you must also provide ranking_params.reranker.text.
  • filters: It supports equivalent filters as the SDK described in Query Filters. For instance: timestamp, entity, keyword, source, topic, sentiment, reporting_entity, etc
  • external_search (optional): Include live web results alongside premium content. Set mode (INCLUDE or ONLY) and values (e.g. ["web"]). For web search, only query.text is used; any other query parameters apply only to Bigdata content. See Search documents.
  • ranking_params: Allow customers to control the search ranking algorithm. The Re-ranker is set by default with a threshold between 0.2 and 0.3 depending of the applied filters. Use ranking_params.reranker.text to provide a custom reranker string when using texts.
  • max_chunks Maximum number of chunks to retrieve. It must be lower than or equal to 1000. The response might contain a lower number of chunks due to discarding duplicated chunks after the matching phase.
The following example searches for news about Pfizer’s partnerships and returns ten chunks.
Output:The response contains an array of documents with the text chunks that matched the query criteria.
The quickest way to test the Search Service is with the Search Playgrounds. It contains a playground per Search Service endpoint.These documentation pages also contain examples:
  • Co-mentions: to see which entities, topics, and sources appear most in your results.
  • Search Volume: to retrieve document and chunk volume statistics over time for a search query, aggregated by date with sentiment analysis.
Need to process large volumes? Batch Search lets you run many requests at once with 50% lower costs.