unitcostai

Embedding Ingestion Cost Calculator

Estimate how corpus refresh volume changes monthly embedding spend, amortized per-user impact, and the price floor for a retrieval-heavy workflow.

How this tool works

This calculator compares baseline and candidate monthly embedding refresh plans, then separates the fixed monthly refresh delta from the per-user economics you need for pricing and rollout decisions.

How It Works

  1. Choose the pricing row used for embedding refresh assumptions.
  2. Set baseline and candidate embedding tokens refreshed each month.
  3. Review fixed monthly cost, amortized per-user impact, and total cost share before changing indexing cadence.

Formula

embedding_ingestion_cost_monthly = embedding_ingestion_tokens_monthly * embedding_input_price_per_token

amortized_cost_per_user_month = embedding_ingestion_cost_monthly / monthly_active_users

Assumptions and Units

  • Currency: USD
  • Token unit: token
  • Embedding ingestion is treated as a fixed monthly refresh term
  • Pricing source: daily snapshot in repo, no runtime scraping

Example Scenario

If monthly refresh scope drops from 1.2B to 400M tokens, compare the fixed monthly savings and the smaller per-user impact at the configured active-user count.

Related resources: AI Workflow Cost Calculator, AI Break-even Price Calculator, LLM Model Cost Comparison, RAG Retrieval Cost Calculator, Reranking Cost Calculator, Cache Savings Simulator, Context Window Cost Calculator, RAG vs Long-Context Calculator, RAG Cost Components Explained, How To Choose Chunk Size and Chunk Count, How Much Does an AI Agent Cost?.

Pricing snapshot: 2026-03-15Provider: AnthropicModel: Claude Sonnet 4.6

Step 1 Provider and Model

Choose the pricing row used for monthly embedding refresh assumptions.

Step 2 Quick Mode

Set baseline and candidate monthly refresh scope before tuning the wider workflow.

Research corpora can grow quickly; quantify refresh cost before widening the index pipeline.

Step 3 Advanced Assumptions

Tune the wider workflow only after the monthly refresh numbers look realistic.
Show advanced inputs

Scenario Share URL

Share this link to load these exact assumptions. Pricing uses the published snapshot shown on the page.

Paste into ChatGPT or Claude to discuss this scenario.

Headline metric

Candidate refresh plan lowers fixed monthly cost

The candidate refresh plan changes fixed monthly embedding cost by -$16.00 and adds -$0.0533 per active user at the current scale.

Baseline embedding cost / month

$24.00

Candidate embedding cost / month

$8.00

Amortized delta / user / month

-$0.0533

Break-even delta / user / month

-$0.0533

Totals

Compare the fixed monthly refresh term and its amortized per-user impact.
MetricBaselineCandidateDelta
Embedding tokens / month1,200,000,000400,000,000-800,000,000
Embedding cost / month$24.00$8.00-$16.00
Embedding share of modeled cost3.1%1.0%-2.0%
Amortized embedding cost / user / month$0.08$0.0267-$0.0533

Component Breakdown

Full workflow context helps show whether refresh is material or just noise.
ComponentBaselineCandidateDelta
GenerationModel input/output token spend for requests.$0.726$0.726$0
RetrievalExtra model input spend from retrieved context chunks.$0.3861$0.3861$0
RerankingReranker cost based on docs scored per request.$1.65$1.65$0
Embeddings IngestionAmortized per-user share of the fixed monthly corpus embedding refresh cost.$0.08$0.0267-$0.0533
Vector DbVector database query cost across all requests.$0.0015$0.0015$0
CacheSavings from cache hits. Negative means lower total cost.$-0.2808$-0.2808$0
InfraNon-model infra overhead per request.$0.044$0.044$0
Sensitivity RankingBaseline sensitivity: cost change when one variable is increased by 10%.
VariableDelta cost %
Requests Per User MonthUser activity level per month.9.7%
Rerank DocsDocs reranked per request.5.7%
Output TokensGenerated tokens per request.1.5%
Retrieved ChunksRetrieved chunk count per request.1.3%
Tokens Per ChunkAverage chunk size in tokens.1.3%
Cache Hit RateFraction of requests served by cache.-1.1%
Input TokensPrompt-side tokens per request.1.0%
Monthly Active UsersActive-user estimate used to amortize fixed monthly embedding refresh.-0.3%
Vector Queries Per RequestVector query count per request.0.0%

Assumptions and Units

Explicit assumptions to keep refresh-cost planning reproducible and auditable.
  • CurrencyUSD
  • Token unittoken
  • Pricing snapshot2026-03-15
  • Selected model rowAnthropic / Claude Sonnet 4.6
  • Monthly users300 active users used to amortize the fixed monthly refresh delta
  • Fixed monthly termEmbedding refresh is modeled as a fixed monthly cost
  • AmortizationPer-user impact divides the fixed monthly delta by active users

Recommended Next Step

Use this section to turn the refresh-cost estimate into an indexing plan.

If refresh cost is material, review infra and indexing constraints before widening your update cadence.

Refresh planning references

RAG Cost Components ExplainedHow To Choose Chunk Size and Chunk Count

Compare infra providers

View Infra Recommendations

Sources and Snapshot

Pricing comes from a daily snapshot generated by batch workflows.

Active Pricing Row

Active pricing row

Anthropic / Claude Sonnet 4.6

  • Input tokens$3 / 1M
  • Output tokens$15 / 1M

Shared retrieval defaults

  • Embedding input$0.02 / 1M
  • Rerank docs$1 / 1K

Continue Analysis

Move to the next tool or guide without losing your current scenario.

Switch tools

Read guides