Qdrant Vector Search Engine
Choosing an Open Source Vector Database? Start Free on Qdrant Cloud
Qdrant is an open source vector search engine under Apache 2.0 that keeps results complete under tight filters. Start a free cluster on Qdrant Cloud with no credit card.
Open source vector search
Why Teams Choose Qdrant for Open Source Vector Search
Every capability below ships in the open source engine and on Qdrant Cloud.
Get complete results under tight filters.
Qdrant applies filters during the graph search and picks a strategy by how selective each filter is, so a narrow tenant or category filter still returns a full result set. Index the fields you filter on.
Run keyword and vector search in one query.
Index the text in your payload, score it with BM25, and fuse it with dense vectors in a single Query API call. Available in Qdrant 1.20 and later.
Update your data in place.
Update points, vectors, and payloads in place. Qdrant merges and reindexes segments in the background as changes accumulate.
Choose what sits in memory.
Keep quantized vectors in RAM and the originals on disk, or move vectors, the index, and payload to disk, collection by collection.
Deployment
Run the Same Engine on Your Servers or on Qdrant Cloud
The engine behind Qdrant Cloud is the open source engine you can download. Pick where it runs today, and move your collections later as your needs change.
Self-host with Docker or Kubernetes.
Run Qdrant with Docker, Docker Compose, or on Kubernetes. Turn on API keys and TLS before production.
Let Qdrant Cloud run it for you.
The same engine as a managed service on AWS, GCP, or Azure, with a free tier, zero-downtime rolling upgrades on multi-node clusters, and automatic shard rebalancing.
Build locally, then ship to Cloud.
Develop against Qdrant in Docker, then point the same client code at a Qdrant Cloud cluster URL and API key.
My AskAI standardized on Qdrant Cloud for its AI support agents after running hybrid search on separate AWS infrastructure. Scaling is two clicks in the dashboard, and the agents deflect about 75% of support requests. Read the My AskAI case study
Migration
Bring Your Self-Hosted Collections to Qdrant Cloud
Start on self-hosted Qdrant and move to Qdrant Cloud when you want managed operations. The Qdrant migration tool copies a collection from one Qdrant instance to another and creates the target collection from the source's configuration.
Keep writing during the move.
Resume an interrupted migration.
Change collection settings as you move.
Reconfigure replication or quantization as part of the migration.
Aracor started on self-hosted Qdrant. Moving to Qdrant Cloud streamlined its operations and freed the technical team for feature development. Read the Aracor case studyStart free with Qdrant Cloud
Getting started
Your First Query on Qdrant Cloud
Create a free cluster.
Sign up for Qdrant Cloud and create a free tier cluster, with no credit card.
Add your vectors.
Create a collection and upload vectors from your embedding model.
Run your first query.
Search by similarity to get the closest points and their payloads.
Common Questions
How do I get started with Qdrant Cloud?
Is Qdrant really open source?
Can I use Qdrant for free?
Can I move from self-hosted Qdrant to Qdrant Cloud later?
Does open source Qdrant have every feature of Qdrant Cloud?
