Qdrant

Qdrant is an open-source vector search engine built for high-dimensional vectors and large scale workflows

Website: qdrant.tech

Qdrant is a fast, open-source vector search engine for building RAG applications and semantic search. Phoenix helps you trace and evaluate your Qdrant-powered applications to understand how well your vector searches are working.

Quick Start

1. Run Qdrant with Docker

docker run -p 6333:6333 qdrant/qdrant

2. Install the Python client

pip install qdrant-client phoenix

3. Basic usage with Phoenix tracing

import phoenix as px
from qdrant_client import QdrantClient
from phoenix.otel import register

# Start Phoenix
px.launch_app()

# Set up tracing
tracer_provider = register(project_name="qdrant-app")
tracer = tracer_provider.get_tracer(__name__)

# Connect to Qdrant
client = QdrantClient(host="localhost", port=6333)

def search_documents(query_vector):
    """Search for similar documents"""
    with tracer.start_as_current_span("search_documents") as span:
        results = client.query_points(
            collection_name="my_docs",
            query=query_vector,
            limit=5
        ).points
        span.set_attribute("result_count", len(results))
        return results

Examples

Complete Qdrant + LlamaIndex + Phoenix Tutorial

Full tutorial showing dense and hybrid search with Qdrant, LlamaIndex, and Phoenix evaluation.

LangChain Qdrant Example

Simple LangChain + Qdrant example with Phoenix tracing.

Further Reading

• Qdrant Docs