All open source
Open SourceRAG5k

Marqo

End-to-end vector search engine with built-in embedding generation.

Marqo is an open-source, end-to-end vector search engine that handles embedding inference and storage in one system, so you index raw text and images rather than precomputed vectors. It supports multimodal search, tensor-based retrieval, and horizontal scaling, removing the need to wire a separate embedding pipeline in front of a bare vector store. It is notable for turning the whole 'embed then store then query' pipeline into a single API.

Repository

marqo-ai/marqo

Language

Python

What you'd build with it

  • Standing up text and image semantic search without managing a separate embedding model and inference service
  • Powering multimodal RAG retrieval where queries and documents can be text or images
  • Replacing a hand-rolled embed-and-index pipeline with one deployable search service

Tags

vector-searchmultimodalembeddings