ChromaDB Has a Free API: The Simplest Vector Database for AI Prototyping

#chromadb #vectordatabase #ai #python

ChromaDB is the simplest vector database — pip install, 4 lines of code, and you have a working semantic search. It runs embedded (in-process) or as a server, with a clean Python and JavaScript API.

Why ChromaDB?

4 lines to start — simplest API of any vector database
Embedded mode — runs in-process, no server needed
Auto-embedding — built-in OpenAI, Sentence Transformers, etc.
Multi-modal — text, images, documents
Free — open source, Apache 2.0

Install

pip install chromadb

Minimal Example (4 Lines!)

import chromadb

client = chromadb.Client()
collection = client.create_collection("docs")

collection.add(
    documents=["AI is transforming search", "Vector databases store embeddings", "RAG combines retrieval and generation"],
    ids=["doc1", "doc2", "doc3"],
)

results = collection.query(query_texts=["how does AI search work?"], n_results=2)
print(results['documents'])  # Most relevant documents

With Metadata Filtering

import chromadb

client = chromadb.PersistentClient(path="./chroma_data")
collection = client.get_or_create_collection("articles")

# Add with metadata
collection.add(
    documents=[
        "Introduction to machine learning algorithms",
        "Best practices for REST API design",
        "How transformers revolutionized NLP",
        "GraphQL vs REST: A comparison",
    ],
    metadatas=[
        {"category": "AI", "year": 2024},
        {"category": "Backend", "year": 2024},
        {"category": "AI", "year": 2023},
        {"category": "Backend", "year": 2023},
    ],
    ids=["a1", "a2", "a3", "a4"],
)

# Semantic search with metadata filter
results = collection.query(
    query_texts=["neural networks"],
    n_results=3,
    where={"category": "AI"},
)

# Combined filter
results = collection.query(
    query_texts=["API design"],
    n_results=5,
    where={"$and": [{"category": "Backend"}, {"year": {"$gte": 2024}}]},
)

Server Mode + REST API

# Start server
chroma run --host 0.0.0.0 --port 8000

# REST API
curl http://localhost:8000/api/v1/collections

curl -X POST http://localhost:8000/api/v1/collections \
  -H 'Content-Type: application/json' \
  -d '{"name": "my_collection"}'

# Connect from Python
client = chromadb.HttpClient(host="localhost", port=8000)

JavaScript

import { ChromaClient } from 'chromadb';

const client = new ChromaClient();
const collection = await client.getOrCreateCollection({ name: 'docs' });

await collection.add({
  documents: ['AI is amazing', 'Vectors power search'],
  ids: ['id1', 'id2'],
});

const results = await collection.query({
  queryTexts: ['artificial intelligence'],
  nResults: 2,
});
console.log(results.documents);

RAG with ChromaDB + OpenAI

import chromadb
from openai import OpenAI

chroma = chromadb.PersistentClient(path="./rag_data")
collection = chroma.get_or_create_collection("knowledge")
openai = OpenAI()

def add_knowledge(texts: list[str], ids: list[str]):
    collection.add(documents=texts, ids=ids)

def ask(question: str) -> str:
    # Retrieve
    results = collection.query(query_texts=[question], n_results=3)
    context = "\n".join(results['documents'][0])

    # Generate
    response = openai.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": f"Answer using this context:\n{context}"},
            {"role": "user", "content": question},
        ],
    )
    return response.choices[0].message.content

Key Features

Feature	Details
Modes	Embedded, client-server, cloud
Embedding	Auto (OpenAI, ST, Cohere) or BYO
Filtering	Metadata + document
Persistence	In-memory or disk
Languages	Python, JavaScript, Ruby, Go
Multi-modal	Text, images

Resources

Building AI apps? Check my Apify actors or email spinov001@gmail.com.

DEV Community