Qdrant

Qdrant is an open-source vector search engine with a managed cloud offering. One source reports the engine is licensed under Apache-2.0 and free to self-host with no feature gating, though I did not check the repository itself. The cloud has a permanent free tier, described by several sources as a single-node cluster with 1 GB of RAM, 0.5 vCPU and 4 GB of disk, with one aggregator saying it needs no credit card and has no time limit. Paid pricing is inconsistent across sources: one lists paid clusters from $0.014 an hour for the smallest production node, another gives paid plans from $25 a month, a third estimates a standard range of $30 to $200 a month with hourly billing, and a fourth says the paid tier has no published rate so you need a calculator or sales contact.

I did not reach qdrant.tech, and the sources run from March to September 2026, so check the official pricing page and calculator. My search found nothing on Qdrant's funding, so I have not included any. Self-hosting is the cost-saving route, since the engine is written for efficient resource use, but you take on operations. Comparisons like Weaviate versus Qdrant pricing are common reading when shortlisting, though each comparison reflects the author's assumptions about vector counts and replication, so test with your own data.

Qdrant suits teams that want a performant open-source vector database with a generous free cluster. Compare Pinecone, Weaviate and Chroma. Browse the MLOps & Infrastructure tools directory for more.

Vector search for RAG

Open source.

Low-cost prototyping

Free cluster.

Is there a free tier?

Yes, a 1 GB cluster.

Licence?

Apache-2.0 per one source.

Paid starting price?

From about $25 a month or $0.014 an hour.

Can I self-host?

Yes.