Chroma

Chroma is a vector database built from the ground up for AI applications. It combines multiple search capabilities—vector embeddings for semantic similarity, full-text lexical search, regex patterns, and metadata filtering—all in one system optimized for retrieving context for large language models and AI agents.

Highlights

• Multi-modal search: vector, BM25/SPLADE lexical, full-text, regex, and structured metadata filtering in a single query
• Open source with high adoption: Apache 2.0 license, 26,000+ GitHub stars, 11 million monthly downloads
• Flexible deployment: managed cloud with pay-as-you-go pricing, free tier with $5 credits, or self-hosted open source
• Developer-friendly APIs in Python, TypeScript, and Rust for easy integration into AI pipelines
• Zero-ops infrastructure with automatic scaling, dataset versioning, and forking for experiment tracking

Chroma is for ML engineers and developers building production AI systems that need efficient, scalable retrieval. The free tier covers prototyping; production use scales with your needs.