Vector DB Pricing Calculator

Estimate monthly cost across the popular vector DB options: Pinecone Serverless, Qdrant Cloud, Weaviate Cloud, and self-hosted pgvector / Qdrant on a small VPS. Inputs are vector count, dimensions, and monthly read/write volume.

How to use the Vector DB Pricing Calculator

Set vector count, dimensions (1536 is OpenAI 3-small / Cohere; 768 is many open-weight; 3072 is OpenAI 3-large), monthly query and write volumes. The estimator runs each vendor's posted pricing formula. Always verify against the vendor's actual calculator before committing.

What drives vector database cost

Vector databases price along a few axes: how many vectors you store, their dimension (which sets the bytes each one occupies), and how many reads and writes you run per month. Managed services like Pinecone, Qdrant Cloud, and Weaviate Cloud each apply their own formula over those numbers, while self-hosting pgvector or Qdrant on a VPS trades that per-unit pricing for a flat server cost you operate yourself.

Because the formulas differ, the cheapest option flips depending on your scale and traffic shape. This estimator runs the same workload — vector count, dimension, query and write volume — through each vendor's model so you can compare like for like. To work out the storage size first, use the embedding storage calculator.

Common use cases

  • Vendor comparison — see how Pinecone, Qdrant, and Weaviate price the same index.
  • Managed vs self-hosted — weigh per-unit cloud pricing against a flat VPS bill.
  • Scale planning — find where one option overtakes another as vectors grow.
  • Traffic modeling — see how query and write volume move the monthly total.
  • Budget estimates — get a ballpark before committing to a vector store.

Frequently asked questions

What inputs drive the estimate?
Vector count, dimension, and monthly query and write volume. Together those set storage size and the read/write charges each vendor bills for.

Which dimension should I enter?
1536 for OpenAI text-embedding-3-small and ada-002, 3072 for 3-large, and around 768 for many open-weight models. Match it to the embedding model you use.

Are these figures exact?
No — they approximate each vendor's posted formula, and pricing shifts over time. Confirm against the vendor's own calculator before committing.

How do I size storage before pricing it?
Use the embedding storage and RAM calculator to get the index footprint, then bring that into the cost comparison here.
Embed this tool on your site

Free to embed, no attribution required (but appreciated). Paste this where you want the tool to appear: