Otter vs Rev vs Descript vs free local transcription: cost per hour compared

An hour of automatic transcription costs between $0.31 and $2.55 on the popular subscription services once you divide the monthly fee by the minutes you actually use, $0.15 to $1.44 through the cloud APIs those services are built on, and about a cent of electricity when the model runs on your own computer. Human transcription is a different product at $119 an hour. The rest of this post shows the working, with prices as published on 14 September 2026.

The short table

“Cost per hour” for subscriptions assumes you use the whole allowance. Almost nobody does, which is the first thing the marketing pages do not mention.

Otter

Otter’s free Basic plan gives 300 transcription minutes a month, caps any single conversation at 30 minutes, and allows three audio or video file imports for the lifetime of the account. Pro is $16.99 a month, or $8.33 a month on an annual plan, for 1,200 minutes, a 90-minute cap per conversation and ten file imports a month. Business is $30 a month, or $19.99 annually, with unlimited minutes and meetings up to four hours.

The number to watch is file imports, not minutes. Otter is built around live meeting capture. If your work is recordings you already have, Pro lets you upload ten of them a month. A podcaster with weekly episodes and a couple of interviews is already at the limit, and the 1,200 minutes are then mostly unused. At ten one-hour files a month on the annual plan, the real cost is $0.83 an hour; on the monthly plan it is $1.70.

Rev

Rev has moved from pay-per-minute AI to seats. The free plan includes 45 AI transcription and caption minutes a month, English only. Essentials is $25.49 per seat per month on an annual plan ($29.99 monthly) with 5,000 AI minutes per user; Pro is $47.99 per seat ($59.99 monthly) with 10,000 minutes. Human transcription is $1.99 a minute with a 99%-plus accuracy promise and delivery within twelve hours.

Five thousand minutes is 83 hours. At full use Essentials is the cheapest subscription here at $0.31 an hour. A freelancer transcribing ten hours a month pays $2.55 an hour, which is more than any cloud API charges and more than Descript. Rev’s pricing rewards volume and punishes light use.

The human tier is worth separating out. At $119.40 an hour it is not competing with software. It is for court, medical and broadcast work where someone has to be accountable for every word, and where the cost of a wrong name is higher than the cost of the transcript.

Descript

Descript sells editing time, not minutes, and the free plan includes 60 media minutes a month. Hobbyist is $16 a month billed annually ($24 monthly) for 10 hours; Creator is $24 annually ($35 monthly) for 30 hours; Business is $50 annually ($65 monthly) for 40 hours.

Per hour that is $1.60, $0.80 and $1.25 on the annual plans. The Creator plan is the one to compare, because 30 hours is a realistic ceiling for a working podcaster. The catch is that media hours count the file you import, not the transcript you export, so a two-hour raw recording that becomes a 40-minute episode costs two hours of allowance.

The APIs underneath

Every one of these products runs speech recognition that costs them fractions of a cent per minute. The published prices for the same capability, bought directly, are $0.006 a minute from OpenAI’s Whisper API, $0.0043 from Deepgram’s Nova-3 batch tier, from $0.0025 at AssemblyAI, and $0.024 at AWS Transcribe. That is $0.15 to $1.44 an hour, with three of the four under 40 cents.

You need to write code to use them, and the audio still leaves your machine, so they are not a consumer option. But they set the floor. When a subscription works out at $2.55 an hour, you are paying $2.20 for the interface, the storage, the summaries, and the company’s margin.

Local transcription

The same open-source model family that powers several of these services, OpenAI’s Whisper, runs inside a modern browser using your graphics card. There is no per-minute price because there is no server. The costs that remain are a one-time download of about 200 MB for the model, which the browser caches, and electricity.

The electricity figure is small. On the desktop we test on, an hour of audio takes about 40 minutes of compute at roughly 60 W of extra draw, which is 0.04 kWh, or about one cent at European domestic rates. A laptop is slower and draws less, so it comes out similar.

What you give up is real, and it belongs in the comparison:

• It needs a desktop Chrome or Edge with a working graphics adapter. Phones and Firefox are not there yet.
• It is English only at the moment.
• The base model is less accurate than the large models the paid services run, particularly on names, technical vocabulary and heavy accents. Read the output before you rely on it.
• It runs at about 1.5 times real time on a desktop. A one-hour recording takes about 40 minutes while the tab stays open. A cloud service returns the same hour in a few minutes.

For notes, searchable archives, first-draft subtitles and anything you would rather not upload, those trade-offs are usually fine. For a deposition they are not.

Three people, three answers

A student recording four one-hour lectures a month. Otter’s free plan cannot take the files (three imports, ever). Descript’s free hour covers one lecture. Rev’s free 45 minutes covers none. Local transcription covers all four for nothing, and the recordings never leave the laptop.

A podcaster with 30 hours of raw audio a month. Descript Creator at $24 a month is the natural fit, because the editing is the product and transcription rides along. Local transcription saves the $24 but not the editor.

A researcher with 100 hours of interviews a year and a consent form that promises no third parties. Every subscription service is a third party. The choice is a local model or a human transcriber under a data agreement, and the difference is roughly $12,000.

Frequently asked questions

Is free local transcription really free? Yes, in the sense that there is no fee, no account and no minute cap. The site is paid for by advertising. Your cost is a model download and some electricity.

Which service is most accurate? All the paid services use large models and score within a few points of each other on clean English audio. The gap between them is smaller than the gap between clean audio and a noisy room. Human transcription is the only option that reliably gets names right.

Do the free tiers stay free? Otter, Rev and Descript have each changed their free allowances more than once in the last few years. Treat any free tier as a trial rather than a plan.

Can I use the cloud APIs without coding? Not directly. They are built for developers. Several small desktop apps wrap them, but then you are back to paying per minute plus the app.

Sources, checked 14 September 2026

• Otter.ai pricing page: otter.ai/pricing
• Rev pricing page: rev.com/pricing
• Descript pricing page: descript.com/pricing
• OpenAI, Deepgram, AssemblyAI and AWS per-minute rates as listed on their pricing pages, collected in our research notes on the same date

Prices change. If you find one that has moved, tell us and we will update the table.