GPT API Cost Calculator
Estimate OpenAI GPT API costs by token volume, request count, and model.
Pricing dataset updated Oct 12, 2026
Estimated monthly cost
$10.00
- Cost per request
- $0.0002
- Input tokens
- $3.00
- Output tokens
- $7.00
- Annualised
- $120.00
Estimates use public list pricing per 1M tokens. Batch, enterprise, and regional rates may differ.
How it works
Start with the average input and output token counts in your API usage logs. Input includes instructions, conversation history, retrieved passages, and tool definitions. Output is billed at a separate rate; which side dominates depends on both the token mix and the selected model.
For a support assistant, sample both short questions and longer conversations before choosing an average. Multiply the estimated per-request cost by all model calls, including retries and follow-up calls, not just customer conversations.
Worked example
Budgeting a support assistant
Consider 50,000 monthly calls averaging 1,200 input and 350 output tokens. That is 60 million input and 17.5 million output tokens. At illustrative rates of $0.25 input and $2 output per million, input costs $15 and output $35, giving $50 per month. These rates are assumptions, not a current model quote. If each conversation triggers two calls, the estimate becomes $100 before any cache savings or extra charges.
Provider references: OpenAI prompt caching