Skip to content
ENExplore Calculators

Batch Inference Cost Calculator

Work out batch inference cost instantly with clear inputs, formula shown and shareable results.

About Batch Inference Cost Calculator

Offline scoring cost is simply total tokens times the blended price, but batch endpoints typically halve that in exchange for a delivery window measured in hours. Cost per item is the figure to compare against the business value of each scored record — at fractions of a cent per item, the bottleneck is usually engineering time rather than inference spend.

Formula

Batch cost

total tokens = items x tokens per item; cost = tokens x price / 1,000,000 x (1 - batch discount)

Worked example

With the default values below, Batch Inference Cost Calculator returns Total batch inference cost of $120.00. Change any input above and every figure here is recalculated in your browser.

Inputs used

Items to processminimum 1
500,000
Tokens per item (input + output)minimum 1
800
Blended price per million tokensminimum 0
0.6
Batch API discountaccepts 0–90 %
50 %

Results

Total batch inference cost
$120.00
Tokens processed
400,000,000
Cost per item
$0.000240
Saved versus real-time pricing
$120.00

Step by step

  1. 1Values used
  2. 2Batch cost
  3. 3Total batch inference cost= 120.00
  4. 4Tokens processed= 400,000,000
  5. 5Cost per item= 0.000240
  6. 6Saved versus real-time pricing= 120.00

Assumptions and limits

  • The result is derived only from the 4 inputs listed above. Anything not among them is not part of the calculation.
  • In particular, this calculator does not take inflation, tax, fees or charges or compounding frequency into account.
  • Figures are rounded for display to 6 decimal places, so totals built from several rounded parts can differ slightly from the unrounded arithmetic.

How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.

Frequently Asked Questions

When is the batch API not suitable?

Anything user-facing. Batch jobs have completion windows of up to 24 hours and no latency guarantee, so they suit enrichment, backfills and evaluation runs.

How do I reduce tokens per item?

Trim boilerplate from the prompt, cap the output with a schema or a max-token limit, and truncate retrieved context to what the task genuinely needs.

Related Calculators

Batch Inference Cost Calculator | Calcrivo