Descript Creator's 30 media hours vs Sume STT at $0.01 a minute

Thirty hours of transcription on Sume costs $18.00, against $24 a month for Descript Creator's 30 media hours. The numbers are close, but they buy different things: Descript is an editor with transcripts built in, and Sume is a set of API jobs that return a transcript, captions or a cut file.

Descript's side comes from its pricing page, read on 2026-10-04. It lists Hobbyist at $16 a month for 10 media hours and Creator at $24 a month for 30 media hours.

Sume's per-minute side

Sume's speech-to-text is $0.01 per audio minute, with a limit of 10 minutes per request. Video inspect runs transcription when you ask for it, at the same $0.01 per audio minute. Thirty hours is 1,800 minutes, so $18.00, split over at least 180 requests of 10 minutes.

What you do not get

Sume does not give you a text-based editor where deleting a sentence removes the audio. A transcript is data you read, and edits are separate jobs such as video trim and video captions. If a person edits by reading and deleting words, Descript's plan is the product you are comparing against, and the hours figure is the wrong measure of value.

The API fits when transcripts feed something else: a caption job, a search index, a quote card, a clip list. Then you pay for minutes actually transcribed and nothing for the hours you did not use.

Which to pick

• A person edits episodes by hand every week: Descript Creator.
• A pipeline transcribes whatever lands in a bucket: Sume, at $0.01 a minute.
• Both: transcribe on Sume for the data, keep the editor for the human pass.