Search breakout topics using hybrid vector + keyword search.
Supports three search modes:
Documentation Index
Fetch the complete documentation index at: https://docs.upriver.ai/llms.txt
Use this file to discover all available pages before exploring further.
Request for searching breakout topics.
Search query text
1Filter by vertical (e.g., 'sports')
Search mode: vector, keyword, hybrid
vector, keyword, hybrid Return raw topics or deduped derived story surfaces
topic, story Maximum number of results. Requesting more than 20 results requires a credits-based plan.
1 <= x <= 100Optional expansions to include in the response. Allowed values: citations, entities. Example: ["citations", "entities"]
Deprecated: use include=["citations","entities"] instead.
Optional source-aware context for query rewriting. Useful for terse external prompts such as Kalshi markets.
Optional Kalshi event URL or ticker. When provided, the server resolves event metadata and uses it for query planning. Mutually exclusive with query_context.
500Enable enhanced relevance scoring for search results. When true, results are scored and re-ranked by how well they match the query intent.
Successful Response
Response for searching breakout topics.
Search mode used
Whether search returned raw topics or derived story surfaces
topic, story Matching topics when surface_mode=topic
Matching story surfaces when surface_mode=story