Semantic ranking
Ranks by meaning, not just keyword overlap. A query about 'resetting credentials' matches 'change your password' even with no shared words.
CatalogLatency Typically 0.5–3 seconds depending on document count Async false Rate Limit Per API key Max Input Up to a few hundred documents per request
cn8/text-search
Send a query and a list of documents, and get them back ranked by semantic
Learn more about Semantic Search Text NLPsyncInstant previewtextanalysisnlp
Try it
Query
DocumentsOne document per line — the query is matched against these.
Result
Results show up here
The sample input is ready — just hit Run.
Overview
Send a query and a list of documents, and get them back ranked by semantic similarity to the query. Each result includes its rank, the original document text, a similarity score, and its index in your input list. The response also echoes the query and the total number of documents ranked. Because matching is by meaning, a query about "resetting credentials" still matches "change your password". Responses are synchronous — one request, one ranked list.
Key capabilities
Rank, score and original index
Each result includes its 1-based rank, the document text, a similarity score, and the document's original index in your input list.
Query echo
The response repeats the query and total document count — handy for logging.
Multilingual
Query and documents can be in different languages — they're mapped into the same meaning space.
When to use it
Document retrieval
Retrieve the most relevant chunks from a document set for a user query.
FAQ matching
Match a user question to the closest FAQ entry by meaning.
Recommendation
Find similar articles, products, or entries from a query or item description.
Input & output
input
A query string and a documents array of strings.
JSON body
output
Ranked results (rank, text, score, index), the echoed query, and the total document count.
JSON
Guides & tips
How it works
- The query and every document are compared by meaning, and the documents are returned ranked from most to least relevant.
- Each result carries its
rank, the documenttext, a similarityscore, and theindexit had in your input list. - All documents are ranked and returned. If you only want the top matches, take the first N entries of
resultsin your own code.
Tips for better results
- Keep documents focused (e.g. paragraphs, not entire articles) for more precise matching.
- For large-scale or repeated retrieval over thousands of documents, pre-compute vectors with Text Embeddings and store them in a vector database.
- Use this endpoint for ad-hoc ranking over modest lists per request.
Specs
Schema
Request body
queryrequiredstringThe search query.
documentsrequiredarray<string>The list of candidate documents to rank against the query.
Response
results[].rankinteger1-based rank position.
results[].textstringThe original document text.
results[].scorenumberSimilarity to the query, from 0 to 1.
results[].indexintegerThe document's original 0-based index in your documents array.
querystringThe query you sent, echoed back.
total_documentsintegerNumber of documents that were ranked.
Pricing
Billed per request.
| Service | Unit | Price |
|---|---|---|
| Semantic Search | item | $0.004/request |
Price is per request, regardless of how many documents you rank.