What does it cost to run a RAG chatbot over 100,000 documents?
Figuring out how to cost a RAG chatbot that answers over 100,000 documents is one of the first budgeting questions AI teams hit. Built for teams building document Q&A, this tool skips the generic rate card and gives you a working estimate you can tune to your traffic. The result reacts live as you adjust the inputs.
Start here: RAG Cost Calculator
Results update automatically as you type.
- Total chunks
- 2,000
- Ingestion (one-time)
- $0.0205
- Vector storage
- 0.01 GB
- Storage / month
- $0.00343
- Per query
- $0.0113
- Queries / month(1,000/day)
- $337.52
- Total / month
- $337.53
Then: RAG Query Cost Calculator
Results update automatically as you type.
- Query embedding
- $8.00e-7
- Retrieved context
- 2,560 tokens
- Generation cost
- $0.00762
- Per query
- $0.00763
- Per month
- $457.55
Why this isn't trivial
The part people underestimate: one-time embedding of 100,000 docs, ongoing vector storage, and per-query retrieval each add cost in different ways. In practice the biggest savings come from right-sizing chunk size and how many chunks you retrieve per query, so it is worth modelling before you commit.
How it's calculated
We estimate this by adding the one-time embedding cost, monthly vector storage, and the per-query generation cost across your query volume. Every figure uses the current provider prices baked into the site (reviewed daily), and you can override any input to match your own assumptions.
Frequently asked questions
Is embedding a one-time cost?+
Mostly — you pay to embed the corpus once, plus re-embedding when documents change.
What makes RAG queries expensive?+
Stuffing many retrieved chunks into each prompt inflates input tokens; retrieving fewer, better chunks is cheaper.
Are these prices up to date?+
Yes. The model prices behind this calculator are refreshed and reviewed daily, so your estimate reflects current provider rates rather than a stale snapshot.
Related
Estimates for planning. Pricing data last reviewed 28 July 2026.