chromadb

What it is and what it does

Chroma is a vector database designed to store and search document embeddings with built-in support for metadata filtering and full-text search. It provides a simple four-function API for creating collections, adding documents with automatic tokenization and embedding, and querying by semantic similarity. The package handles the infrastructure layer for AI applications that need to retrieve contextually relevant documents.

The library supports both in-memory prototyping and persistent storage modes, with a client-server architecture available via the command line. It depends on 28 runtime packages including pydantic, onnxruntime, grpcio, and opentelemetry. The project is actively maintained with regular releases and 29058 GitHub stars, though the substantial dependency footprint and two known vulnerabilities warrant attention before production deployment.

Use it for:

• Build retrieval-augmented generation systems that fetch relevant documents to augment prompts.
• Implement semantic search over document collections with metadata-based filtering.
• Prototype vector search applications locally before scaling to a hosted service.
• Store and query embeddings from custom embedding models with optional full-text search.
• Add similarity-based recommendation or deduplication logic to data pipelines.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Chroma is a vector database and search infrastructure that stores, indexes, and queries document embeddings with optional metadata filtering and full-text search capabilities.

Yes, with conditions. Chroma is worth installing for vector search applications where you need a managed embedding store with a simple API. Active maintenance, permissive license, and top-5000 popularity are strong signals. However, 28 runtime dependencies create medium install friction, and two known vulnerabilities (GHSA-f4j7-r4q5-qw2c, PYSEC-2026-311) require review before production use. For prototyping, install freely; for production, verify vulnerability impact and assess dependency footprint.

Install

chromadb on PyPI

pip install chromadb

uv add chromadb

poetry add chromadb

Installing chromadb

Before you install

Medium install friction due to 28 runtime dependencies including onnxruntime, grpcio, and opentelemetry packages. Active maintenance with recent commits and regular Monday releases; last release was 101 days ago. Requires Python 3.9 or later.

License in practice

Apache 2.0 permissive license allows commercial use, modification, and distribution with minimal restrictions—suitable for most production and proprietary projects.

Quickstart

pip install chromadb

import chromadb
client = chromadb.Client()
collection = client.create_collection("my-docs")
collection.add(documents=["doc1", "doc2"], ids=["id1", "id2"])
results = collection.query(query_texts=["search query"], n_results=2)

Requires Python 3.9 or later; onnxruntime and grpcio dependencies may require system libraries on some platforms.

Verify before relying

• Whether the 28 runtime dependencies create significant bloat or startup-time overhead in typical usage patterns.
• Performance characteristics and scalability limits for in-memory versus persistent/server modes.
• Details on the two known security vulnerabilities (GHSA-f4j7-r4q5-qw2c, PYSEC-2026-311) and their impact.

Package facts

Evidence: chromadb-1.5.9-cp39-abi3-macosx_10_12_x86_64.whl; chromadb-1.5.9-cp39-abi3-macosx_11_0_arm64.whl; chromadb-1.5.9-cp39-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; chromadb-1.5.9-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; chromadb-1.5.9-cp39-abi3-win_amd64.whl

Tags

More Database packages