Term
Chroma
Chroma (ChromaDB) is an open-source vector database for storing and searching embeddings. It is widely used in RAG systems to retrieve relevant text passages via similarity search and feed them to language models.
Chroma — explained in detail
Chroma, often referred to as ChromaDB, is an open-source vector database. Its purpose is to store embeddings — numeric vector representations of text, images or other data — together with the underlying content and metadata, and to make them searchable by similarity.
Technically, Chroma uses the HNSW algorithm (Hierarchical Navigable Small World) for indexing. It builds a graph in which similar vectors are connected as neighbours, enabling search in roughly logarithmic rather than linear time. For local, persistent storage Chroma relies on SQLite by default; for production workloads a client-server mode is available.
Chroma can be accessed through Python and JavaScript clients and integrates with common LLM frameworks such as LangChain and LlamaIndex. By default it generates embeddings via Sentence Transformers, but it can also use models from OpenAI, Cohere or custom embedding functions. A 2025 rewrite of its core in Rust removed bottlenecks caused by Python’s Global Interpreter Lock and noticeably improved throughput for both writes and queries.
Example / Practical use
A typical use case is Retrieval Augmented Generation (RAG): a knowledge base — such as manuals or support articles — is split into chunks, each chunk is converted into an embedding and stored in Chroma. When a user asks a question, that question is also encoded as a vector, the most similar chunks are retrieved from Chroma and passed to the language model as context. The model then answers based on concrete sources rather than on its training knowledge alone.
Distinction from related terms
Chroma is a standalone database designed specifically for vector search. By contrast, pgvector is an extension for the relational database PostgreSQL that embeds vector capabilities into an existing SQL database. Chroma is often quicker to get started with for prototypes and smaller projects, whereas pgvector excels when vector data needs to be tightly coupled with existing relational data. Both work with the same embeddings but differ in architecture and operating model.
Entdecke mehr
Saving Tokens with Claude: 6 Principles That Make Experts Twice as Fast
How I turned my CLAUDE.md from a style guide into a token budget — 6 principles for lower cost, less waiting, and more honest reporting.
GlossarMilvus
Milvus is an open-source, distributed vector database for fast similarity search over large sets of embeddings. It is licensed under Apache 2.0 within the LF AI & Data Foundation; the main developer is the company Zilliz.
LexikonVector Databases Compared
What vector databases do, when you need one, and how Chroma, Weaviate, Milvus, Qdrant and pgvector stack up against each other.