Workload assumptions

Live FX rates come from Frankfurter. If conversion is unavailable for a request, the calculator falls back to the model source currency and marks that row.

Models to compare

Leave nothing selected and the page falls back to a default trio that already has live snapshots.

Quick token estimate

Optional helper for rough sizing before you set request token numbers.

Reset

Current workload summary

Selected models 1
Display currency USD
Monthly request volume 300,000
Request shape 2,000 in · 1,000 out
Cache and batch 20% cache · Batch off
Budget Not set
FX mode Source currency
From English words 0
From Chinese chars 0
From pages 0
Estimated total tokens 0

Estimate only. Default assumptions: 1 token ~= 0.75 English words, 1 token ~= 1.5 Chinese characters, 1 page ~= 500 English words.

How the calculator interprets inputs

  • `Cache hit ratio` discounts only the cached share of input tokens.
  • `Batch discount` applies only when the stored snapshot lists a batch ratio for that model.
  • `Budget fit` means the maximum monthly requests you can afford with the exact request shape above.
Model Monthly cost 1k requests Blend / 1M Cache savings / month Batch savings / month Budget fit
Gemini 2.5 Flash
gemini-2.5-flash
Gemini
Updated Mar 30, 02:54
Verified Mar 30, 02:54
Gemini Developer
Currency USD
standard input=$0.30 (text / image / video) | $1.00 (audio); standard output=$2.50; cache=$0.03 (text / image / video) | $0.1 (audio) | $1.00 / 1,000,000 tokens per hour (storage price)
Cache saves $32.40 No batch savings
$897.60 $2.992 $0.85 $32.40 $0.00
Set budget to see fit

Estimate only. Actual billing may differ by tokenizer behavior, cache hit rate, and provider rules.