Vector Database Cost Calculator
Estimate the monthly cost of a managed vector database from stored vectors, dimensions and query volume — storage plus per-query charges combined.
How it works
Size (GB) = vectors × dims × bytes × replicas ÷ 1e9 · Cost = size × storage rate + (queries ÷ 1,000) × query price
Managed vector databases bill on two axes: how much you store and how much you search. Storage scales with vectors × dimensions × precision, multiplied by replicas since each replica holds the full set. Query cost scales with request volume. For read-heavy retrieval workloads the query line usually dominates, so caching frequent queries and trimming replicas matters more than shaving vector size — while for huge, rarely queried corpora, storage and quantization are where the savings live.
Worked example
Five million 768-dim float32 vectors on 1 replica occupy 5,000,000 × 768 × 4 ÷ 1e9 ≈ 15.4 GB. At $0.25/GB-month that is $3.84 storage. One million queries at $0.05 per 1,000 adds $50. Total ≈ $53.84 per month.
Frequently asked questions
Why do replicas multiply storage cost?
Each replica is a full copy of the index kept for redundancy or extra query throughput, so two replicas store — and are billed for — twice the data. Add replicas only when you need the availability or read capacity.
Storage or queries — which usually costs more?
For interactive search apps, query volume often dominates; for large archival collections that are searched infrequently, storage does. This calculator splits the two so you can see which lever matters for your workload.
How can I cut the bill?
Quantize vectors (float16 or int8) to shrink storage, cache or batch repeated queries to cut request counts, and prune stale vectors. Fewer replicas also help if your availability target allows it.
Related calculators
Educational estimate — real costs and memory depend on your provider, framework, model architecture and pricing, which change often. Verify against your own bills and profiling before committing budget.