Qdrant
High-performance vector database with filtering, payload storage, and a REST/gRPC API — built for production AI.
Qdrant
Qdrant is an open-source vector database and similarity search engine written in Rust. It stores embedding vectors alongside JSON payloads, enabling filtered nearest-neighbor search at scale. It is a popular choice as the retrieval backend for RAG pipelines and agent memory systems.
Key Features
- Filtered search — combine ANN search with arbitrary JSON payload filters in a single query
- Named vectors — store multiple embedding spaces per record (dense + sparse + ColBERT)
- Quantization — scalar and product quantization with on-the-fly rescoring
- Snapshots — point-in-time collection snapshots for backup and migration
- Distributed — horizontal sharding and replication built-in
- REST + gRPC + Web UI — full API surface with an interactive dashboard
- MCP server — official
qdrant-mcpallows agents to store and retrieve memories as vectors
Quick Start
docker run -p 6333:6333 qdrant/qdrant
from qdrant_client import QdrantClient
from qdrant_client.models import Distance, VectorParams, PointStruct
client = QdrantClient(":memory:")
client.create_collection("docs", vectors_config=VectorParams(size=384, distance=Distance.COSINE))
client.upsert("docs", points=[PointStruct(id=1, vector=[0.1]*384, payload={"text": "hello"})])
result = client.search("docs", query_vector=[0.1]*384, limit=3)
print(result)
Install via ai-supply
npx ai-supply add qdrant-vector-store
Curated mirror of the open-source Qdrant project (Apache-2.0). Install upstream from the repository.
Compromise signals — malicious or tampered code (leaked secrets, backdoors, a dropped executable) — reduce the score, and known dependency CVEs carry a bounded penalty (they warrant review but never QUARANTINE — update the dependency to clear). Other dangerous-by-capability traits are risk surface, expected for some capabilities. Every finding is mapped to its OWASP control below.
Findings mapped to the OWASP Top 10 for LLM Applications (2025) and the OWASP Machine Learning Security Top 10. Expand any flagged control for the exact findings — compromise reduces the score; expected/risk-surface do not, except a known CVE, which carries a small bounded penalty (high/critical → Review).
The same gate an agent runs before installing (POST /api/v1/trust/qdrant-vector-store/check). Click a policy:
Consume Qdrant programmatically. Authenticate with an API key or session — see Authorize an agent.
# Agents: CHECK BEFORE YOU INSTALL (no auth) — score, grade, level, capability manifest
curl https://ai-supply.store/api/v1/trust/qdrant-vector-store
# Gate against your org policy (returns { pass, violations })
curl -X POST https://ai-supply.store/api/v1/trust/qdrant-vector-store/check \
-H "Content-Type: application/json" \
-d '{"minGrade":"B","denyPermissions":["shell"],"denyUnknownEgress":true}'
# CLI
npx ai-supply add qdrant-vector-store
# REST (install → download)
curl -X POST https://ai-supply.store/api/v1/listings/qdrant-vector-store/install \
-H "Authorization: Bearer $AIM_KEY"
# MCP tool
install_listing({ "slug": "qdrant-vector-store" })OpenAPI spec →Curated mirror — latest upstream source. See the repository for tagged releases.