Both turn speech into text — but one keeps every 'um' and false start, and the other tidies it into clean, readable copy. Picking the right style up front saves rework, cost and time. Here's how to choose.
In this article
When you order transcription, the first question a good provider asks is: verbatim or clean-read? It sounds like a technicality, but it changes the finished document completely — and the right answer depends entirely on what you'll use the text for.
Verbatim transcription
Verbatim captures exactly what was said, word for word. That includes filler words (“um”, “you know”), repetitions and false starts. In strict or full verbatim it also notes non-verbal detail — laughter, long pauses, [inaudible] markers and background sounds. Nothing is cleaned up, because in the right context those details are the data.
Use verbatim for legal proceedings and depositions, academic and market research where speech patterns are analysed, medical records, and any situation where exact phrasing, hesitation or tone could matter as evidence.
Clean-read (intelligent) transcription
Clean-read — also called intelligent or clean verbatim — keeps the full meaning but removes the noise: fillers, stutters, and repetition that add nothing. Light grammar tidying makes it flow. The result is text a reader can move through quickly without losing anything of substance.
Use clean-read for interviews and podcastsyou'll publish, content destined for subtitles, corporate video, marketing, and anything meant to be read rather than analysed. It is the most common choice by far.
How it affects cost and turnaround
Verbatim generally takes longer to produce — capturing and formatting every utterance is more work — so it usually costs a little more and takes a little longer than clean-read for the same audio. Deciding up front avoids paying for detail you don't need, or ordering a clean-read when a court or research protocol actually requires verbatim.
Not sure? Tell us the use
The simplest way to choose is to tell us what the transcript is for, and we'll recommend the style. We transcribe audio and video in 100+ languages, verbatim or clean-read, and can translate the transcript or turn it into subtitles in the same pass. Send us your files for a quick, itemised quote.