Key facts
| Managed option | Plugsky RAG collections with automatic chunking, embedding and citations |
| Retrieval modes | Keyword, vector and hybrid search with optional reranking |
| Standalone vectors | POST /v1/embeddings works with Weaviate, Qdrant or any store |
| Citations | Every query returns ranked chunks with source attribution |
| Data handling | Per-collection encryption at rest; API data is not used to train models |
| Deployment | Managed, VPC, on-prem and air-gapped options |
| Free tier | plugsky-micro and plugsky-lite, no card required |
| Product status | Live |
TL;DR
- Weaviate leans toward a platform: modules, tenancy and its own query language.
- Qdrant leans toward a focused engine: filtering depth and predictable operations.
- Both support hybrid retrieval, which matters for exact terms and identifiers.
- Compare the surrounding features and operational cost, not just search speed.
- A managed RAG service removes the store choice when retrieval is not the product.
How it works, step by step
- List the features you need beyond vector search: tenancy, filters, modules, APIs.
- Prototype both with the same corpus, embedding model and question set.
- Test filtered search, hybrid weighting and latency at realistic concurrency.
- Check multi-tenant isolation if different customers share one deployment.
- Review upgrade, backup and monitoring requirements for each engine.
- Decide whether to run the store or use managed collections, then document the exit path.
Try it yourself
Open the vector database comparison →
Where Weaviate fits
Weaviate is an open-source vector database that behaves like a platform. It offers REST, GraphQL and gRPC interfaces, built-in hybrid search that combines keyword and vector scoring, multi-tenancy for shared deployments, and a module system that can connect vectorisers and generative capabilities to collections.
That breadth shortens some projects: hybrid search and tenancy are configuration rather than custom code. The cost is more surface area to learn and operate, and a query model that differs from standard SQL or REST conventions, which teams should weigh against their familiarity and tooling.
Where Qdrant fits
Qdrant is an open-source engine written in Rust, focused on vector search with a typed payload filtering system, dense and sparse vectors, configurable indexes and quantisation options for memory control. It exposes a clean API for collections, points and filters, and can run self-hosted or as a managed cloud service.
Its appeal is a smaller, predictable operational footprint: fewer concepts to learn, straightforward scaling, and filtering powerful enough for multi-tenant applications. Teams that want an engine rather than a platform often prefer that focus.
How to compare them honestly
Run the same corpus, embeddings and queries through both. Measure recall at your target k, filtered-search latency, hybrid search quality on queries that mix meaning with identifiers, and the effort to delete or update documents. Multi-tenant filtering deserves its own test, since that is where query complexity grows.
Then compare everything around search: backup and restore, upgrades, monitoring, client library maturity and the operational burden your team can absorb. Those factors usually decide long-term satisfaction more than a small latency difference.
The managed path with Plugsky
If owning a vector database is not the goal, Plugsky RAG collections provide managed ingestion, automatic chunking and embedding, keyword, vector and hybrid retrieval, optional reranking and citations, with private deployment options for stricter requirements. If you prefer your own store, POST /v1/embeddings works standalone and returns vectors you can index anywhere.
Compare architectures with the vector database comparison, then start free with plugsky-micro and plugsky-lite or the 14-day full-access trial. Current plans are on the live pricing page.
Honest comparison
| Factor | Weaviate | Qdrant | Plugsky RAG collections |
|---|---|---|---|
| Design focus | Platform with modules and tenancy | Lean search engine with deep filtering | Managed retrieval end to end |
| Interfaces | REST, GraphQL and gRPC | REST and gRPC | OpenAI-compatible REST |
| Hybrid search | Built in | Dense and sparse vectors | Keyword, vector and hybrid built in |
| Operations | Self-host or managed cloud | Self-host or managed cloud | Managed with private deployment options |
| Best for | Feature-rich platform needs | Focused, filter-heavy workloads | Teams that want answers, not a store |
Frequently asked questions
Are Weaviate and Qdrant both open source?
Yes. Both can be self-hosted, and both also offer managed cloud services for teams that prefer not to operate them.
Which has better hybrid search?
Both support keyword and vector retrieval together. Test with your own queries, especially ones mixing exact identifiers with natural language, before deciding.
Can I use Plugsky with either engine?
Yes. POST /v1/embeddings is standalone, so you can generate vectors with Plugsky and store them in Weaviate, Qdrant or another database.
What about multi-tenancy?
Weaviate offers tenant isolation as a platform feature, while Qdrant handles multi-tenancy through payload filters. Both approaches work but differ in operational shape.
Is there a free plan?
Yes. The free plan includes plugsky-micro and plugsky-lite with 2 API keys and no credit card, and a 14-day full-access trial is available.
How is pricing structured?
Self-serve plans are flat monthly with unlimited fair-use usage and no per-token charges or overage fees. See the live pricing page for current plans.