Chroma

Founded by Alexander Kvamme in 2022

Store, search, and retrieve vector embeddings and text data for applications

Pricing model

Free Tier, Paid

Rating

★ People love it

Time to value

Quick Setup (< 1 hour)

You can use Chroma to store, manage, and query high-dimensional vector embeddings alongside text and metadata. It connects to your applications using Python or JavaScript libraries, letting you execute fast semantic searches, sparse keyword lookups with BM25, and regex queries over large datasets. You can run it locally during development or deploy it in cloud environments with automated data tiering to object storage. It also supports collection branching for testing, metadata filtering, and background synchronization to index external documentation and code repositories.

What Chroma does

Index vector embeddings and text chunksQuery collections using nearest neighbor algorithmsFilter search results by metadata tagsExecute full-text and regular expression searchesFork collections to test alternative retrieval setupsSync external web pages and code repositories into collectionsSparse vector search supporting BM25 and SPLADESemantic similarity search over high-dimensional vectorsFull-text search with trigram and regex matchingMetadata filtering and faceted query capabilitiesDataset forking for A/B testing and experimentationAutomatic data tiering between memory caches and object storageNative client libraries for Python and JavaScript