Estimate the cost of embedding a document corpus, ongoing re-embed cost, and the raw vector storage footprint.
OpenAI1,536 dimensions$0.020 / 1M tokens
100,000 docs × 300 tokens = 30,000,000 tokens per full embed pass
One-Time Corpus Embed
Full pass over the corpus
$0.6000
Ongoing Monthly Cost
At 1× re-embed / month
$0.6000
Raw Vector Storage
1536-dim float32, 100,000 vectors
585.9 MB
Storage figure is the raw vector bytes only — most vector databases add index overhead (HNSW graphs, metadata, replication) on top of this, often 1.5–3× the raw size.
About the Embedding Cost Calculator
Embedding a large document corpus for RAG has two cost components people underestimate: the one-time embed pass over the full corpus, and the recurring cost of re-embedding as documents change. This calculator projects both, plus the raw vector storage footprint your chosen model's dimensionality implies — before you commit to a vector database sizing decision.
Key Features
Covers OpenAI text-embedding-3, ada-002, Gemini, Cohere embed-v3, and Voyage-2.
One-time embed cost, monthly re-embed cost, and raw vector storage size (float32).