Chroma is open-source data infrastructure for AI applications, centered on storing and retrieving information for embedding-based and language-model workflows. Its ecosystem supports vector search and is designed to make it straightforward for developers to add retrieval and memory capabilities to AI applications.

The project exposes a deliberately small core API while its hosted Chroma Cloud service adds serverless vector, hybrid, and full-text search. This allows developers to start with a simple local or application-oriented interface and move toward managed infrastructure when scale, operational simplicity, or additional retrieval modes become important.

Chroma is widely applicable to retrieval-augmented generation, semantic search, agent memory, recommendation-style retrieval, and other systems that need to find relevant context from large collections of data. It is particularly attractive to AI developers who want a purpose-built retrieval layer without adopting a traditional database workflow first.