Embedding Cost & Vector Storage Calculator
Estimate text embedding cost, compare reviewed Batch rates, and size raw float32 or float16 vectors locally.
Ready. Enter workload totals; nothing is uploaded.
Dimensions source · Pricing source · Reviewed 2026-10-15. USD estimates exclude taxes, free tiers, media and account-specific discounts. No embeddings are generated.
Enable JavaScript to calculate locally.
Price tokens; store vectors.
Embedding cost uses total source tokens and the selected dated USD rate. Raw storage uses vector count × dimensions × bytes per component. One million tokens do not necessarily produce one million vectors: choose the vector count from your chunking plan.
A concrete storage example.
10,000 vectors × 1,536 dimensions × 4 bytes occupy 61,440,000 raw bytes, or 61.44 MB in decimal SI units. float16 halves raw bytes but can change numerical precision. Indexing, metadata, replication and database overhead are additional; no database storage price is inferred.
Reviewed choices, explicit limits.
This text-only calculator includes OpenAI text-embedding-3-small/large and Gemini Embedding 2. Dimensions are a conservative reviewed subset. It does not generate vectors, invoke a provider or account for model-specific per-request token limits, media, free tiers or discounts. Split token workloads into valid requests separately. Rates are observed from October 15, 2026; earlier dates have no invented history.