Choose your product

Compare the two products and choose the right starting point.

Choose by	Embedded retrievalZvec	Workspace searchZvec-Grep
Typical scenarios	Store and query embeddings generated by your application Index parsed records with text, vectors, and scalar metadata Control schemas, indexes, filters, and hybrid queries in code	Search a code or document workspace as it already exists Let Zvec-Grep scan, parse, chunk, embed, and index the files Retrieve workspace evidence from the CLI or an AI agent
Use through	Native SDKs	CLI, MCP, or AI agents
Choose it when	You own data parsing and vector generation and need a programmable retrieval engine	You want ready-to-use workspace search without implementing ingestion or retrieval
Data flow	Index SchemaParsed records + vectorsinsert / upsertIndexes Search Vector / text / filtersquery()Ranked documents	Index Workspace filesScan + parse + chunkBM25 + vector index Search QueryExact / hybridCLI / MCPHuman / agent