qdrant
A production vector search engine that ships as a single binary and filters like a real database.
What it is
Qdrant is a vector similarity search engine written in Rust. You store vectors with a JSON payload attached to each one, then run queries that combine nearest-neighbour search with conditions on that payload — the closest items to this vector that are also in stock and under a given price, for example.
Unlike the container-stack databases, Qdrant ships as one self-contained binary with nothing else to install: no metadata store, no object storage. That makes it the easiest of the production-grade vector databases to get running, while still supporting sharding, replication and persistence to disk.
It suits teams building semantic search, recommendation or RAG systems that have outgrown an in-process library but do not want to operate a cluster. The filtering support is what separates it from a bare nearest-neighbour index.
Why we recommend it
Payload filtering is a first-class feature
Filtering on metadata alongside the vector search is what turns a similarity demo into a usable product. Many engines bolt this on as an afterthought; here it is part of the query language.
One binary, no dependencies to babysit
You can be running a persistent vector database in a couple of minutes, and upgrading means replacing one file. That matters when there is no platform team to maintain a cluster.
Rust means steady tail latency
No garbage collector pauses, so latency stays predictable under load — and the tail latency is the number users actually notice on a search page.
Quantization keeps memory under control
Scalar and product quantization let you trade a little recall for a large cut in RAM use, which is often the difference between fitting on one machine and needing several.
How it scores
These are our editorial ratings, not benchmark results. We score on how the project behaves in practice — how long it takes to get running, how good the documentation is, how responsive the maintainers are, and how well it holds up over time. The licence and star count above are read from GitHub and are a snapshot.
Pros and cons
Strengths
- Single self-contained binary — no external metadata store or object storage
- Rich filtering that combines payload conditions with vector similarity in one query
- Quantization and multiple distance metrics built in
- Excellent, example-heavy documentation
- Apache 2.0 with an active maintainer team and a managed cloud option
Trade-offs
- Distributed mode is more hands-on than the single-node experience and less mature
- No built-in text embedding — you bring vectors from a separate model
- A smaller ecosystem of third-party integrations than the older databases in this category
- The default single-node setup has no high availability, so a critical deployment needs more work
How to get it
Direct download
Clicking a button downloads the file straight from the project's release — you stay on this page.
One-line install
Prefer the terminal? Copy a command and you are done.
docker run -p 6333:6333 qdrant/qdrant
Source and documentation
At a glance
| Category | Vector DB |
|---|---|
| Licence | Apache 2.0 |
| Written in | Rust |
| GitHub stars | 34,476 |
| Repository | qdrant/qdrant |
Licence and star count were read from the GitHub API when this page was generated and are a snapshot. Check the repository for the current figure and the authoritative licence text.
Other Vector DB tools
We are not affiliated with qdrant or its maintainers. We list it because it is genuinely useful and genuinely free. If you plan to build a commercial product on top of it, read the licence on the repository first.