Loading page
About this tool
Turn per-million-token pricing into cost per request, per 1,000 and per million, with retries, caching and a tier comparison.
This calculator converts per-million-token AI pricing into the number engineers actually budget with: cost per 1,000 requests. The formula is (input tokens x input rate + output tokens x output rate) / 1,000,000 per request, adjusted for the share of input served from a prompt cache and for billed retries — then compared side by side across five capability tiers from frontier to nano so you can see what stepping down a model tier saves.
Open AI Cost Per 1000 Requests Calculator on AltFTool — it loads instantly in your browser.
Enter the values you already know.
Fine-tune the options to match your scenario.
Read the result and use it in your planning or reporting.
Turns dollars-per-million-token quotes into per-request, per-1,000 and per-million costs instantly.
Failed attempts are still billed and cached input is cheaper — both are in the maths, not a footnote.
Shows the identical workload priced across representative frontier to nano rate bands.
Multiply your average input tokens by the input rate and output tokens by the output rate, divide by one million, then multiply by 1,000. Example: 1,200 input and 350 output tokens at $0.15 and $0.60 per million tokens costs $0.00039 per request, which is $0.39 per 1,000 requests.
Because request volume is the number teams actually forecast — a support bot handles conversations, not tokens. Once token sizes per request are measured, cost per 1,000 requests maps directly onto traffic projections and per-user unit economics, which per-token rates do not.
Yes. Timeouts discovered after the model responded, rate-limit retries, refusals and schema-validation failures all bill the tokens of the failed attempt, so a 5% retry rate raises the whole bill by 5%. This calculator applies the retry percentage as a multiplier on every cost component.
Output tokens typically cost 4 to 5 times more than input tokens per unit — $15 versus $3 per million is a common ratio. If your requests generate long completions, the output line can be most of the bill even when prompts are much larger than responses; capping response length and trimming verbosity are the quickest savings.
Add the AI Cost Per 1000 Requests Calculatorwidget to your blog or website — free, responsive, no signup. Just keep the “Widget by AltFTool” credit link visible.
<iframe src="https://www.altftool.com/embed/widget/ai-cost-per-1000-requests-calculator"
title="AI Cost Per 1000 Requests Calculator — free AltFTool widget"
width="100%" height="640" style="border:0;border-radius:12px;overflow:hidden"
loading="lazy" referrerpolicy="no-referrer-when-downgrade"></iframe>
<p style="font-size:12px;margin:4px 0 0">Widget by <a href="https://www.altftool.com/tools/all/ai-cost-per-1000-requests-calculator?utm_source=embed&utm_medium=widget">AltFTool — free online tools</a></p>