Rev or Dalm: words a person checks, or audio no one else hears

Published · Updated

Rev and Dalm both turn a recording into a transcript, and both write meeting summaries — Rev’s notetaker by joining the call as a bot. With Rev you upload a file or record on your phone, and you can order a transcript made by people, who then work with your file. Dalm has no transcribers and no file upload: the speech is recognised on your Mac and the audio goes nowhere; the transcript syncs to our servers automatically, and a language model reads it to write the summary. Choose Rev when a person must check every word; Dalm when no one else should hear the recording.

Rev and Dalm, side by side

RevDalm
Who turns speech into textIts speech models, or — if you order it — its human transcribers, who work with your fileA speech model on your Mac; no person is involved
Getting a recording inUpload from your device, Dropbox, Box, Google Drive or OneDrive, or paste a link; or record in its iPhone and Android appA file on your Mac, dragged onto the window or opened with ⌘O; or a recording made in the app
The audio afterwardsYou upload it, and Rev works on itNever sent anywhere; deleted once the text exists, with the one exception below; a file you bring in stays where it was
The text afterwardsIn your Rev accountSynced to our servers automatically; no setting keeps it on the Mac only
A call or a meetingIts notetaker bot joins Zoom, Google Meet, Microsoft Teams and Webex from your calendar or a link; recording a call from the computer with nothing joining it: Not foundYour microphone and what your Mac plays, as two streams, with nothing joining the call; a meeting in the room through the Mac’s microphone
While it recordsA live transcript, in the phone app and from the notetakerA draft of what has been said, marked preliminary; the full transcript when you stop
Free minutes45 a month of AI transcription, English onlyNo limit on recording minutes on any plan
Fixing the textEdit the transcript in its editor; export to Word or PDFCorrect any line, with the original a click away; documents as Markdown, Word or PDF
Typing by voice into other appsNot found; in its help a dictation is a recording with spoken instructions for the transcriberOne key, into any app; the text syncs to our servers
AppsWeb, iPhone, AndroidMac with Apple Silicon · macOS 15.0 or later; iPhone app in development

The audio is deleted once the text exists. Not found means Rev’s documentation does not describe it, not that it cannot be done. What Rev does is taken from its pricing page, terms, product pages and help center, checked on .

Where Rev is better than Dalm

  • A person checks the words. You can order a transcript made by Rev’s transcribers, at $1.99 a minute, and subscribers get a discount. Dalm has no transcribers: the audio never leaves your Mac, so no one could listen and correct it. What you get is the model’s transcript, which you fix yourself; that text syncs to our servers.
  • Phones, Android and the browser. Rev works in a web browser, and its iPhone and Android apps record, show a live transcript as you go and send a recording for machine or human transcription. Dalm is a Mac app for Apple Silicon · macOS 15.0 or later; its iPhone app is in development.
  • Files from wherever they live. Rev pulls a file from Dropbox, Box, Google Drive or OneDrive, or from a link. Dalm takes a file from your Mac — dragged onto the window or opened with ⌘O.
  • Connections to other tools. Rev connects to Clio, Dropbox, Box, Google Drive and OneDrive. Dalm connects to Google Calendar.
  • A notetaker that goes without you. Rev’s bot joins Zoom, Google Meet, Microsoft Teams and Webex meetings from your calendar, on your behalf, and writes a summary with decisions and action items. Dalm records from the Mac in front of you: nothing joins the call, so nothing is recorded when you are not there.
  • Captions and subtitles made for you. Rev sells captions made by people and subtitles in other languages.
  • Cheaper for a small group. Rev Essentials, for up to three users, is $29.99 a month for each user, or $25.49 billed annually; Dalm Team is $40 a month for each member, or $32 a month billed annually (paid plans are not open yet).

Where Dalm is better than Rev

  • No one else hears the recording. At Rev, a file sent for human transcription is opened by its transcribers, and its terms say uploaded content is analysed by its models and may be used to train them. In Dalm the speech is recognised on your Mac and no audio is sent anywhere. What travels is text: transcripts and dictations go to our servers automatically, and a language model reads each meeting’s transcript to write its summary. What leaves your Mac, and what does not.
  • Minutes that do not run out. Rev’s free plan gives 45 minutes of AI transcription a month, and its paid plans 5,000 or 10,000 for each user. No plan puts a limit on recording minutes, on the length of a recording, or on how long anything is kept.
  • A file you already have, with nothing to upload. Dalm reads any audio or video file macOS can read, up to four hours long, with no limit on size, on the Free plan, and leaves the original where it was. Turning an audio file into text without uploading it is a guide of its own.
  • Calls and rooms with nothing in the call. Your microphone and the sound your Mac plays arrive as two streams, and when Zoom, Microsoft Teams, Slack or FaceTime takes the microphone, a prompt asks “Record this meeting?”. Rev’s meetings come through its notetaker bot; recording a call from the computer without one is not described in its help.
  • A summary that starts from your notes. What you type during a meeting decides what the summary leads with, and each point keeps the words behind it. Every quote shows you where it was said — open it and the transcript opens at that place.
  • Less to pay on your own. Dalm Pro is $16 a month, or $12 a month billed annually; Rev Essentials is $29.99 a month, or $25.49 billed annually.

Paying by the minute or by the month

RevDalm
Free45 minutes of AI transcription a month, one user, English onlyA plan, not a trial
Billed monthlyEssentials $29.99 a month for each user; Pro $59.99Pro $16 a month; Team $40 a month for each member
Billed annuallyEssentials $25.49 a month for each user ($305.90 a year); Pro $47.99 ($575.88 a year)Pro $12 a month; Team $32 a month for each member
AI minutes on paid plans5,000 a month for each user on Essentials, 10,000 on Pro; no cap on a custom-priced planSee below
A transcript made by people$1.99 a minute, with a discount for subscribersNot offered
A machine transcript without a planFrom $0.25 a minute, as its help listsNot sold by the minute

On the Dalm side the Free plan has no end date, and it takes files you bring in. No plan puts a limit on recording minutes, on the length of a recording, or on how long anything is kept. What the paid plans add is on the pricing page, and how a group shares one space has a page of its own. Until Dalm is released, the waitlist is the way in. Rev’s prices are its regular ones, from its own pricing page, checked on .

From file to transcript: who handles the recording in each

StepRevDalm
Getting it inYou upload a file, link one from cloud storage, record on a phone, or let the notetaker join the callYou open a file on your Mac, or record a call or a room from it
Turning speech into textIts speech models; if you order a human transcript, its transcribers can open your file as they workA model on your Mac; no audio is sent anywhere
Keeping itIts help: data is kept as long as needed to provide its services or as required by lawAudio deleted once the text exists (the exception is under the first table); the text syncs to our servers
The summaryIts notetaker writes one for meetings it joinedA language model from an outside provider writes one after each meeting; it reads the transcript, your notes and the names you gave voices — never the audio
Training modelsIts terms: content may be used for continuous training of its speech recognition and other AI models, not for generative AI. Its help: recordings customers submit train its speech recognition, large language models and third parties are not trained on them, and enterprise customers and its top plan can opt outDalm does not train models of its own; what the provider of its language model does with the text is not set out on this site

If a recording should be heard by no one but you — a client, a candidate, a source — Dalm is built for that, and the price is that no person can check the words. If every word has to be checked by a person, Rev does that, and then a person hears the recording.

Which one fits which job

  • Rev, if a transcript must be checked by a person, if you record on Android or work in a browser, if your files already live in Dropbox, Box, Google Drive or OneDrive, or if you want a notetaker to attend meetings you cannot.
  • Dalm, if you work on a Mac with calls, meetings in the room and files you already have, and the recording should not reach anyone else’s computer; the text still syncs to our servers.
  • Both, if most recordings can stay with you and a few need a person’s check: Dalm for the everyday ones, Rev for those few — knowing that those files then go to Rev.

Rev also runs Temi, an automatic transcription service; on temi.com opened with a page pointing Temi users to Rev.

Comparing more than two? Seven transcription tools, sorted by the job each does best. In general terms, local and cloud recognition, step by step.

Questions

Is Dalm a good Rev transcription alternative?
Dalm is a Rev alternative for people on a Mac who want their recordings heard by no one else: the speech is recognised on the Mac and no audio is sent anywhere. The text does leave: the transcript syncs to our servers, and a language model reads it for the summary. If you need a transcript checked by people, apps for Android or the web, or a notetaker that attends without you, Rev is the stronger choice.
How much does Rev cost?
Rev's free plan gives 45 minutes of AI transcription a month, in English. Essentials is $29.99 a month for each user, or $25.49 billed annually; Pro is $59.99, or $47.99 billed annually; a transcript made by people is $1.99 a minute, and its help lists machine transcripts from $0.25 a minute. Dalm Pro is $16 a month or $12 a month billed annually, Team is $40 a month for each member, and there is a Free plan. No plan puts a limit on recording minutes, on the length of a recording, or on how long anything is kept. These are Rev's regular prices from its pricing page; the date is on this page.
Does Rev use my recordings to train its AI?
Rev's terms say uploaded content is analysed by its speech recognition and other AI models and may be used to train them, but not to train generative AI. Its help says submitted recordings train its speech recognition, not large language models, and that enterprise customers and its top plan can opt out. Dalm does not train models of its own; what the provider of its language model does with the text is not set out on this site. The audio is not sent anywhere, and the transcript syncs to our servers.
Does a person listen to my audio at Rev?
At Rev, its transcribers work with your file when you order human transcription, and its help adds that support and engineering staff have limited visibility into transcripts and files for support and debugging. In Dalm no one listens: the speech is recognised on your Mac, and the transcript syncs to our servers. The audio is deleted once the text exists.
Can Rev record a meeting without a bot?
Rev's meetings are recorded by its notetaker bot, which joins Zoom, Google Meet, Microsoft Teams and Webex from your calendar or a link; its phone app records a meeting in the room. Recording a call from the computer with nothing joining it is not described in Rev's help. Dalm records calls on your Mac with nothing in the call, and meetings in the room through its microphone.
Does Dalm upload my file to transcribe it?
No. Dalm reads an audio or video file where it lies on your Mac, up to four hours long, on the Free plan, and the original is not copied. The speech is recognised on the Mac; the transcript then syncs to our servers. Step-by-step, see how to turn an audio file into text without uploading it.