File transcription

What it does

File transcription turns a recording you already have, such as an interview, a lecture, or a meeting, into text. It works in the desktop app on macOS and Windows with your Dictāta plan. No API key is needed.

Each transcript gets punctuation and paragraphs. Hesitation sounds, stutters, and repeated words are removed; the content itself is not shortened or summarized. Speaker labels and timestamps are not added.

Open File transcription

Open the main app window and select File Transcription in the sidebar.

Drop one or more files on the page, or click it to choose them.

Supported files

• Formats: MP3, WAV, M4A, OGG, MP4, MOV, and WebM.
• Length: up to 3 hours per file.
• Size: up to 1 GB per file. For a larger video, export its audio track (for example as MP3 or M4A) and add that instead.

Video files are handled on your computer: Dictāta reads the sound track and sends only the audio for transcription, never the video picture.

Transcribe several files

Every file you add gets its own card. Files wait as Ready until you start them.

1. Add the files.
2. Click Transcribe (for example, Transcribe 2 files).
3. Follow each file's progress on its card.

Files are transcribed one at a time, in the order you added them. A long recording is processed in parts, so its card shows which part is running, the time elapsed, and an estimate of the time left.

You can keep working while files are processed:

• Add more files with Add more files or by dropping them on the page. Files added during a run join the queue.
• Leave the page, for example to open History. Transcription continues as long as the app stays open.
• Remove a file that has not started with its X button.

Read, copy, and save a transcript

When a file is done, its card shows Saved to History. Click Show text to read the transcript on the page. You can edit it there; your changes are saved to the History entry when you click outside the text box.

Use Copy to put the text on the clipboard, or Download .txt to save it as a text file. Open History takes you to all your saved transcripts.

Clear finished removes finished files from this page only. Their transcripts stay in History.

Stop, continue, and retry

Click Stop to pause the queue. The file that was running stops right away, keeping the parts that were already finished, and files that had not started go back to Ready.

A stopped or failed file keeps the parts that were already transcribed. Click Continue to transcribe only the remaining parts. If a file fails, the other files in the queue still run.

If the app closes before a file is finished, add the same file again within 24 hours: its card shows where it will continue from, and the parts that are already done are not sent again.

Minutes and limits

File transcription uses the same monthly minutes as dictation. Long pauses are shortened before the audio is sent, and the skipped silence does not count toward your minutes.

If your plan's minutes run out during a run, the current file stops and the files waiting after it go back to Ready. You can continue them when minutes are available again. See Account and subscription for plans.

The Usage page shows how your minutes split between dictation and files:

The visuals on this page are drawn from the desktop app's interface; file names and numbers are examples.

Privacy

The audio of each file is sent to Dictāta Cloud for transcription, in parts. Finished transcripts are saved in your local History for the signed-in account.

Troubleshooting

• "This file format cannot be read": convert the file to MP3, WAV, or M4A and add it again.
• "No speech was found in this file": the recording contains only silence or is too quiet to detect speech.
• A file does not appear when added again: a file that is still on the page, and not finished, is not added twice.

See Troubleshooting for general problems.