Skip to main content
It offers a robust search that includes semantic search, vector search, keyword (BM25) search, sparse vector search, multi-vector search, hybrid search, reranking, and more. Moreover, it’s designed to support massive-scale datasets and is accessible via Python SDK or JavaScript SDK.

​
Guides

Semantic Search

Search using natural language — no embedding pipeline required.

Vector Search

Bring your own embeddings and run nearest-neighbor search.

Sparse Vector Search

Efficient high-dimensional retrieval using sparse representations.

Multi-Vector Search

Index and query multiple vector fields per document.

Keyword Search

BM25-based full-text search for term matching across documents.

True Hybrid Search

Combine semantic and keyword signals in a single query.

Reranking

Re-score results with a cross-encoder model for higher precision.

​
APIs

Document API

Upsert, query, get, and delete documents.

Management API

Create, list, get, and delete collections.

​
SDKs

https://mintcdn.com/topk/8NBkS0nek3e9o6Vi/icons/python.svg?fit=max&auto=format&n=8NBkS0nek3e9o6Vi&q=85&s=97cbee7891538170fd752e1afbc98095

Python SDK

Full Python SDK reference.
https://mintcdn.com/topk/8NBkS0nek3e9o6Vi/icons/js.svg?fit=max&auto=format&n=8NBkS0nek3e9o6Vi&q=85&s=7642cf18b45f52a70f141214b3d0eca1

JavaScript SDK

Full TypeScript/JavaScript SDK reference.

​
Digging Deeper

Consistency

Understand TopK’s consistency model and write freshness guarantees.

Multi-Tenancy

Partition data by tenant within a single collection.