GPT-5 Cost Calculator Examples

The easiest way to estimate GPT-5 API costs is to separately calculate input and output usage.

How the estimate is calculated

For the original gpt-5 API model, OpenAI currently lists standard rates of $1.25 per million input tokens and $10 per million output tokens, with cached input priced lower.

The basic calculation is:

What can change the total

Cost = (input tokens ÷ 1,000,000 × input rate) + (output tokens ÷ 1,000,000 × output rate)

Consider a customer-support application handling 100,000 conversations per month.

Plan with a realistic usage example

Suppose each interaction averages:

Input: 1,000 tokens Output: 300 tokens

Practical checks before you proceed

That equals:

100 million input tokens 30 million output tokens

Practical checks before you proceed

Approximate cost:

100 × $1.25 = $125 input cost

Practical checks before you proceed

30 × $10 = $300 output cost

Estimated total = $425 per month

Practical checks before you proceed

Now suppose prompt optimization cuts the average output from 300 to 200 tokens. Monthly output becomes 20 million tokens, reducing the output component to $200.

That saves approximately $100 per month at the same traffic level.

Practical checks before you proceed

Real billing can also involve caching, tools, Batch processing, different processing tiers, and other model-specific charges. Newer OpenAI models have their own pricing, so always select the exact API model when performing calculations.

A useful AI cost calculator should therefore ask for model, number of requests, average input tokens, average output tokens, cached tokens where applicable, and expected monthly traffic.