# pip install chromadb

import chromadb

# Connect to Chroma Cloud

client = chromadb.CloudClient()

collection = client.get_or_create_collection(name="my_documents")

collection.add(

documents=[

"Hello, world!",

"Chroma is a retrieval database for AI applications.",

"Try vector, full-text, regex, and metadata search."

],

metadatas=[

{"source": "tutorial", "type": "greeting"},

{"source": "tutorial", "type": "description"},

{"source": "tutorial", "type": "features"}

],

ids=["doc1", "doc2", "doc3"]

)

# Vector search for "greeting"

results = collection.query(

query_texts=["greeting"],

n_results=1,

where={"source": "tutorial"},

include=["documents", "metadatas"]

)

print(results)