What vector database should I use for semantic search with filters?

If you want semantic search + metadata filters, the best vector database depends on your scale, latency needs, and how “database-like” you want it to feel. Here’s the practical shortlist:

Best overall choices

1) Pinecone

Best if: you want the easiest managed option with strong performance and filtering.
Pros:

• Very good semantic search performance
• Solid metadata filtering
• Fully managed, low ops burden
• Good for production apps

Cons:

• Can be pricier than self-hosted options
• Less flexible than a general-purpose DB

Use it when: you want reliability and speed without managing infrastructure.

---

2) Weaviate

Best if: you want a feature-rich vector database with strong filtering and hybrid search.
Pros:

• Strong metadata filtering
• Hybrid search support (keyword + vector)
• Open source and managed options
• Good developer experience

Cons:

• Can be heavier operationally if self-hosted
• More knobs to understand

Use it when: you want flexibility and may need hybrid search later.

---

3) Qdrant

Best if: you want fast filtering, open source, and a clean API.
Pros:

• Excellent metadata filtering
• Fast and efficient
• Easy to self-host or use managed
• Good for production search workloads

Cons:

• Less “batteries included” than Weaviate in some areas
• Fewer built-in extras than some competitors

Use it when: filters are important and you want a strong open-source option.

---

4) Postgres + pgvector

Best if: your dataset is moderate and you already use Postgres.
Pros:

• Simple architecture
• SQL filters are excellent
• Great if you want one database for everything
• Easy to integrate into existing systems

Cons:

• Not as scalable/optimized for very large vector workloads
• Performance can lag specialized vector DBs at high scale

Use it when: you want simplicity and your scale is manageable.

---

Quick recommendation by scenario

• Easiest production managed service: Pinecone
• Best open-source all-around: Qdrant
• Best feature-rich open-source with hybrid search: Weaviate
• Best if you already use SQL/Postgres: pgvector

What matters most for filters

For semantic search with filters, make sure the DB supports:

• Fast metadata filtering
• Indexing on filter fields
• Boolean filters like AND, OR, IN, range queries
• Optional hybrid search if you need keyword + semantic

My default pick

If you want a single recommendation:

• Qdrant if you want open source and strong filtering
• Pinecone if you want the simplest managed path

If you want, I can also give you a recommendation based on your:

• dataset size
• number of users / queries per second
• cloud vs self-hosted
• filter complexity
• budget