Chroma

Chroma is the open-source, AI-native vector database built for LLM application developers. It runs in-memory, as a local persistent store, or as a distributed server — all with the same clean Python (and JavaScript) API. No infrastructure expertise required.

Key features

• Simple API — add, query, update, delete in four lines of Python
• Multimodal — store text, images, and arbitrary embedding vectors
• Metadata filtering — combine vector similarity search with structured where-clause filters
• Embedding functions — built-in support for OpenAI, Cohere, Hugging Face, and Instructor embeddings
• Multiple modes — in-process (no server), local persistent, and client-server with Chroma Cloud
• Integrations — LangChain, LlamaIndex, Haystack, CrewAI, AutoGen, and more

Quick start

npx ai-supply add chroma-vector-database

# Or install directly
pip install chromadb

import chromadb

client = chromadb.Client()  # In-memory
# Or: client = chromadb.PersistentClient(path="./chroma_db")

collection = client.create_collection("my_docs")

collection.add(
    documents=["AI agents are transforming software", "Vector search enables semantic retrieval"],
    metadatas=[{"source": "blog"}, {"source": "paper"}],
    ids=["doc1", "doc2"]
)

results = collection.query(
    query_texts=["What is semantic search?"],
    n_results=1
)
print(results["documents"][0])

Curated mirror of the open-source Chroma project (Apache-2.0). Install upstream from the repository.