Chroma

Chroma is an open-source vector database designed for developer simplicity, its primary design goal is making semantic search and retrieval-augmented generation accessible to developers who need to add AI memory to their applications without operational complexity. The API is intentionally minimal: create a collection, add documents with embeddings (or let Chroma embed them using a built-in model), and query by semantic similarity in a handful of lines of Python or JavaScript. Chroma runs in-process for development and single-server setups, or as a persistent client-server deployment for production, no infrastructure management required for most use cases. It generates embeddings automatically using sentence-transformers models, removing the separate embedding step that other vector databases require.

Metadata filtering enables hybrid queries that combine semantic similarity with exact field matching, finding the most relevant documents from a specific date range or category without post-processing the similarity results. Chroma integrates natively with LangChain and LlamaIndex, making it the default vector store choice in many tutorial codebases and quick-start templates. The free open-source version is self-hosted; Chroma Cloud provides a managed version for teams that don't want to operate their own instance. Commonly used for RAG prototyping, chatbot memory systems, and document Q&A applications where developer speed matters more than raw performance at extreme scale.