Cook with AI Enrollment closed
AI & LLMs Runs locally

Embeddings Cost Estimator

Estimate embedding API cost, chunk counts, vector storage size, and rough monthly bills for your corpus.

ELI5

It estimates the cost of turning your documents into numbers an AI can search.

A little more detail

What this tool does

Embeddings convert text into vectors so an application can find related passages. This tool estimates how many chunks and vectors your documents create, plus API and storage costs.

Use it to
  • Budget the first indexing run for a RAG project
  • Estimate vector count and storage size before choosing a database
  • Compare chunk sizes and embedding models
Interactive workspace Results update as you type
Corpus
Chunking
Embedding model

 · 

Re-embed frequency

Numeric inputs sync to the URL. Assumes full corpus re-embedded each period.

Estimate
Chunks per document
Total chunks / vectors
Embedding tokens (initial)
Initial embedding cost
Monthly re-embed cost
Vector storage (float32)

Storage = vectors × dimensions × 4 bytes. Indexes and metadata add overhead not modeled here.

Typical vector store costs

Static rough numbers for storage only — query/read pricing not included.

Everything runs in your browser. Prices are a mid-2026 snapshot — check vendor pages before you budget.

RAG has two bills: embedding API calls and storing/querying vectors. This estimator counts chunks from document size, chunk length, and overlap, then multiplies by embedding price and vector dimensions.

Real pipelines add metadata, hybrid search indexes, and incremental updates. Use this for order-of-magnitude planning — then prototype with a few hundred docs and read your actual dashboard.