RetrievalSearch

Hybrid Search

Vector similarity and keyword matching run together, then merge — so an exact product code is found as reliably as a paraphrase.

Vector search understands meaning and misses identifiers. Keyword search finds INV-2026-0412 and misses the question that never used those words. Running only one of them is a decision to be bad at half your queries.

Ragenta runs both, deduplicates the candidates, and hands the merged set to the reranker. Every answer shows which passages survived and which fell below the cut.

Use Hybrid Search in your own workspace.

Create a workspace, connect a source, and put this to work in a few minutes.

Book a demo →

Related entries

Cohere Rerank 3

Popular

Cross-encoder reranker that reorders retrieved chunks by true relevance before they reach the model's context.

RerankingRetrieval

Voyage 3

Retrieval-tuned embedding model for English corpora where recall on long technical documents matters.

EmbeddingRetrieval

Semantic Chunker

Splits documents on structure rather than character count, so a passage retrieved is a passage that makes sense.

IngestionRetrieval