Rev — AI and human transcription for audio and video

Use when

You need speech turned into text and want the accuracy tier to be a choice rather than a vendor lock—machine transcripts in minutes for volume work, human-verified transcripts with timestamps when a transcript has to hold up, plus caption files in SRT and VTT for video and a speech-to-text API when the job belongs in code.

Skip when

You need meeting notes wired to your calendar invites, live captions with nothing to upload, or an unlimited free allowance for casual personal use—or transcription is only a preamble to editing the recording itself.

Try instead

Use on-site audio utilities when a transcript is not the deliverable: pull the track out of a video first, or trim the recording down before deciding what actually needs transcribing.

• Extract audio
• Cut audio
• Video sound

• RevThis page

  Pricing

  Free tier with a monthly AI-minute allowance, paid subscription tiers that include more AI minutes per seat plus a discount on human work, or pay-as-you-go per audio minute with no plan at all

  Learning cost

  Upload a file or record from the mobile app, then polish in the browser transcript editor; the API path needs a key and an HTTP or SDK call

  Output limits

  Machine transcripts in minutes at roughly 96% class accuracy, human-verified transcripts on a 12-hour standard turnaround, captions and subtitles including non-English, and SRT, VTT, JSON, or plain text output

  Privacy

  Rev states uploads are encrypted, data is not sold or used to train third-party LLMs, and compliance covers HIPAA, CJIS, and SOC 2 Type II depending on the plan; legal and investigative workflows are a first-class part of the product

• Otter.ai

  Pricing

  Free tier with a monthly minute cap, paid per-seat tiers that raise the cap and add admin and integration features

  Learning cost

  Join or upload a meeting and let it transcribe; the product leans on meeting capture rather than file-based orders

  Output limits

  Strong on speaker-labelled meeting summaries and action items; file length and minutes are bounded by the plan tier

  Privacy

  Meeting audio and transcripts are stored in the workspace and governed by its retention and admin settings

• Descript

  Pricing

  Free tier with a monthly transcription allowance, paid tiers that raise limits and unlock higher export quality

  Learning cost

  Import media, then edit the video by editing its transcript—the transcription is a means to cutting, not the end product

  Output limits

  Output is tied to the editing project, so the transcript is a by-product of the cut rather than a deliverable in its own right

  Privacy

  Media and transcripts live in the project workspace, with the publisher’s terms governing retention

1. Rev — official site
2. Rev — pricing
3. Rev AI — speech-to-text API
4. Otter.ai — official site
5. Descript — official site

Free tier

Yes

Pricing summary

Rev stacks three pricing shapes on one account. A free tier covers a limited monthly allowance of AI transcription minutes. Paid subscription tiers raise the AI minute allowance per seat and include a discount on human-verified work. Pay-as-you-go is also available, where machine transcription is billed per audio minute at a low rate and human-verified transcription, captions, and subtitles are billed per minute at a substantially higher one because a person reviews the output. Legal-formatted transcripts are priced by the page and sit outside subscriptions. Per-minute rates change, so confirm current tiers and allowances on the official pricing page—this page does not list currency amounts.

Reviewed on 17 September 2026 · Rev — pricing

• AI transcription

  Upload audio or video and receive a text transcript in minutes, with the publisher citing 96%+ accuracy and automatic punctuation and capitalization.

• Human transcription

  Order a transcript reviewed by a person when accuracy cannot be negotiated, with a standard turnaround measured in hours and timestamping included.

• Captions and subtitles

  Generate caption and subtitle files for video, including non-English captioning and burned-in options, in the caption formats publishing platforms accept.

• Interactive transcript editor

  Correct wording, adjust timing, and rename speakers while the audio or video plays alongside the text so edits stay anchored to the source.

• Order add-ons

  Timestamping, verbatim output, rush turnaround, and legal-formatting options can be added per order, each with its own per-minute or per-page rate.

• Rev AI speech-to-text API

  An asynchronous batch API for pre-recorded files and a streaming API for real-time transcription, with speaker separation, language identification, and JSON, SRT, or VTT output.

• Security and compliance posture

  Rev states that uploads are encrypted, data is not sold or used to train third-party LLMs, and compliance options span HIPAA, CJIS, and SOC 2 Type II depending on plan.

• Mobile recording

  Record from the phone app and let the transcript generate automatically, which suits interviews and field notes captured away from a desk.

1. 1

   Decide machine or human first

   Machine transcription suits volume work where a transcript is a working document; order human review when the transcript is itself the deliverable or has to survive scrutiny.

2. 2

   Upload or record

   Submit audio or video files on the web, or record directly in the mobile app when the source is a live interview or meeting.

3. 3

   Add only the options you need

   Timestamping, verbatim, rush, and legal formatting are per-order add-ons, so costs track the specific requirement instead of a blanket upgrade.

4. 4

   Clean up in the editor

   Fix names and jargon in the transcript editor with playback synced to the text, then export the transcript or move the captions into your video pipeline.

5. 5

   Wire it into your app when it repeats

   If transcription is a recurring step in software rather than a one-off order, call the speech-to-text API and receive results through webhooks instead of handling files by hand.

• Video and content teams

  Produce accessibility captions and subtitle files without a separate captioning subscription.

• Podcast and interview producers

  Turn recorded conversation into transcripts, show notes, and quotes, and reach for human review only on the episodes that need it.

• Developers adding speech features

  Use the batch or streaming speech-to-text API for recorded archives, in-product captions, or bulk processing with webhook callbacks.

• Legal, investigative, and newsroom teams

  Rev’s strongest current focus: evidence review, legal-formatted transcripts, and transcription of recordings where accuracy and confidentiality are the point.

Two accuracy lanes, one account

Machine and human transcription are separate services with different speeds and per-minute rates rather than quality settings on a single one.

Turnaround expectations

AI transcripts arrive within minutes, while human-verified transcription and captions are quoted on a 12-hour standard turnaround, with rush available as a paid add-on.

Human work carries a minimum

Human transcription is billed with a per-order minimum, so very short clips are never the economical path to human review.

Legal output is priced and routed separately

Legal-formatted transcripts are produced by legal-trained transcriptionists, priced per page, and are not part of a subscription.

API and product are distinct surfaces

Rev AI offers asynchronous and streaming speech-to-text with official SDKs, while the main Rev app is built around file upload, recording, and the transcript editor.

Positioning has broadened

Rev’s homepage currently leads with an investigative and legal platform, so expect the surrounding marketing to be framed around evidence review even though general transcription and captions remain core services.