Skip to main content
cost.botzone.ai
Sign in

OpenAI

OpenAI API cost calculator

Price your GPT usage by the month. Enter the calls you make and the tokens each one carries, and the sum rewrites itself line by line, including the part most calculators hide.

Nothing here is gated. No signup, no email, no credentials.

Your usage

Change anything here and every number on the right updates straight away.

Rates come from the same published price table Cost uses to price the calls it meters.

The same calls and the same tokens, priced on a second model.

Start from a typical workload

These load ordinary figures for each shape of work. Change any of them below; the sum follows.

Every request your app makes to the model API, across all routes.

System prompt, retrieved context and the user message. Roughly 750 words is 1,000 tokens.

What the model writes back. Output usually costs about five times as much as input.

25%

The share of your input tokens served from the prompt cache. If the model has no published cached rate, this changes nothing and the working says so.

Estimated model spend

€1,734per month

Exact, in euro
€1,733.63
In dollars
$1,884.38
Per call
$0.012563
Per year
€20,804

The working

Every step, in order. If we quoted you a figure by email, this is how we got it.

  1. 1Model: gpt-4o on OpenAI.
  2. 2Published rate: $2.50 per 1M input tokens, $10.00 per 1M output tokens, $1.25 per 1M cached input tokens.
  3. 3Usage: 150,000 calls a month, 3,000 input tokens and 600 output tokens per call.
  4. 4Cache: 25% of the input is a cache hit, so 750 tokens are billed at the cached rate and 2,250 at the full input rate.
  5. 5Input per call: 2,250 / 1,000,000 x $2.50 = $0.005625
  6. 6Cached input per call: 750 / 1,000,000 x $1.25 = $0.000937
  7. 7Output per call: 600 / 1,000,000 x $10.00 = $0.006000
  8. 8Cost per call: $0.012563
  9. 9Cost per month: 150,000 x $0.012563 = $1,884.38
  10. 10In euro at 0.92 USD to EUR: $1,884.38 x 0.92 = €1,733.63
  11. 11This counts API token cost only. It does not include cache-write charges, retries, embeddings, batch discounts or anything you spend outside the model API.

Converted at a standing rate of 0.92 USD to EUR. The dashboard uses the live daily rate instead, so your real invoice will differ by a percent or two.

What a cheaper model would cost

gpt-4o
€1,733.63
gpt-4o-mini
€104.02

€1,629.61a month less

€19,555 a year

Would quality hold?

On your traffic, nobody knows until it is tested, and an untested switch is a bet against quality. That is usually why the model stays where it is. Cost shadow-runs your real traffic against the cheaper model, scores it with an independent judge across five dimensions, and recommends the switch only once the cheaper model matches on 95% of them.

That is a verified recommendation and a projected saving. It does not switch anything for you. Send a usage export or measure it on the free tier.

The working, compared

The same arithmetic on the second model, and the difference.

  1. 1Compared with: gpt-4o-mini on OpenAI, at exactly the same usage.
  2. 2Published rate: $0.15 per 1M input tokens, $0.60 per 1M output tokens, $0.08 per 1M cached input tokens.
  3. 3Cost per call on gpt-4o-mini: $0.000754
  4. 4Cost per month: 150,000 x $0.000754 = $113.06, which is €104.02.
  5. 5Difference: €1,733.63 - €104.02 = €1,629.61 a month cheaper on gpt-4o-mini.
  6. 6Over twelve months: €1,629.61 x 12 = €19,555.29.
  7. 7Same tokens and the same cache hit rate on both sides. A cheaper model may need longer prompts or more retries, and this does not model output quality.
Show the full sum for gpt-4o-mini
  1. 1Model: gpt-4o-mini on OpenAI.
  2. 2Published rate: $0.15 per 1M input tokens, $0.60 per 1M output tokens, $0.08 per 1M cached input tokens.
  3. 3Usage: 150,000 calls a month, 3,000 input tokens and 600 output tokens per call.
  4. 4Cache: 25% of the input is a cache hit, so 750 tokens are billed at the cached rate and 2,250 at the full input rate.
  5. 5Input per call: 2,250 / 1,000,000 x $0.15 = $0.000338
  6. 6Cached input per call: 750 / 1,000,000 x $0.08 = $0.000056
  7. 7Output per call: 600 / 1,000,000 x $0.60 = $0.000360
  8. 8Cost per call: $0.000754
  9. 9Cost per month: 150,000 x $0.000754 = $113.06
  10. 10In euro at 0.92 USD to EUR: $113.06 x 0.92 = €104.02
  11. 11This counts API token cost only. It does not include cache-write charges, retries, embeddings, batch discounts or anything you spend outside the model API.

Want the full breakdown by route?

We can send you a written report that takes this estimate apart: which model tiers you could drop to, what prompt caching would actually save you, and where the same spend usually hides in a codebase like yours. Leaving an email is entirely optional. The calculator above is complete without it and nothing on this page is hidden behind a form.

We store your address and the figures above so we can write the report. Nothing else, and no third-party trackers. See the privacy page.

How this actually gets billed

The spread between OpenAI's models is wider than the spread inside most other providers. GPT-4o mini prices input at a fraction of GPT-4o, so the single biggest lever on an OpenAI bill is usually which routes genuinely need the larger model rather than any prompt tuning.

The reasoning models bill their thinking. On o1 the tokens the model generates while working are charged as output tokens even though you never see them, so a route that looks cheap by prompt size can be expensive by bill.

Cached input is billed at half the normal input rate on the models that support it. That is a smaller discount than Anthropic's, which changes where caching is worth the engineering.

If you would rather not guess at token counts, upload a usage export on the free analysis page and we will work from your real numbers. Or wrap one client with the SDK and see it measured, on the free tier.

Questions people ask

Does this cover the reasoning models?
It prices o1 and o1-mini at their published rates. Be aware that reasoning tokens are billed as output, so the honest way to use this page for those models is to raise the output token count well above what the visible answer suggests.
Can I compare this against Claude or Gemini?
Yes. Switch the provider in the calculator, or start from the Claude versus GPT comparison page, which opens with both models side by side.
Why euro and not dollars?
Because the bill a European team defends internally is in euro. The page states the conversion rate it used, and the dollar figure stays visible in the working.

Price something else