Loading page
About this tool
Estimate speech-to-text plus AI summarisation cost per audio minute, per hour and per file.
This calculator works out what one minute of recorded audio costs to process end to end, combining speech-to-text billed per audio minute with language-model summarisation billed per token. It converts audio duration into an estimated transcript using your speaking rate, then applies the standard English approximation of roughly three-quarters of a word per token to size the input and output token bills. Useful for podcast teams, support and sales-call analysts, researchers and anyone budgeting a transcription pipeline before turning it on.
Open Cost Per Transcription Minute 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.
Per-audio-minute transcription and per-token summarisation are normalised to a single cost per minute.
Speaking rate turns duration into words and tokens, so you do not need a transcript to estimate token spend.
A share breakdown makes it obvious when human proofing, not the API, is the real cost driver.
Multiply the price per audio minute by 60. At $0.006 per minute that is $0.36 an hour for speech-to-text alone; adding an AI summary of a 9,000-word transcript typically costs well under a cent more at current small-model token prices. Human proofing is usually the largest line item once it is involved.
About 9,000 words at a conversational 150 words per minute. Dictated or slow speech runs nearer 110 words per minute (6,600 words an hour) and fast interviews around 180 (10,800 words an hour), which is why the speaking rate is an input here rather than a fixed assumption.
Divide the word count by about 0.75, since one token averages roughly four characters or three-quarters of an English word. A 9,000-word hour is therefore around 12,000 tokens. Non-English text, code and heavy punctuation tokenise less efficiently, so treat this as an estimate rather than an exact count.
Commonly quoted ratios run from about 2x to 6x real time depending on audio quality and the number of speakers, so a one-hour recording can take two to six hours to clean fully. Enter your own measured minutes per audio hour rather than a default, as this figure dominates the total when a person is paid to do it.
Add the Cost Per Transcription Minute 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/cost-per-transcription-minute-calculator"
title="Cost Per Transcription Minute 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/cost-per-transcription-minute-calculator?utm_source=embed&utm_medium=widget">AltFTool — free online tools</a></p>