Import audio and video

If you already have a recording of a class or a lecture video, you can import it. kk transcribes it into timestamped bilingual captions and, like any session, can generate notes, search it and answer questions about it.

How to import

  1. In the library’s “Audio and video transcription” section, select “Import audio or video” or drag a file in.
  2. In the dialog, confirm the settings: the recognition service, whether to translate at the same time (into your translation language), the translation method, and whether to identify speakers. The dialog shows an estimated time.
  3. Select “Start transcription”. You can cancel at any point while it runs.

You cannot import while a recording is in progress, and only one file is processed at a time; import the next once it finishes.

Supported formats and limits

Item Details
Audio MP3, M4A, WAV, FLAC and more
Video MP4, MOV, MKV and more (the audio track is read)
Size and length Up to 4 GB and 6 hours per file

If a format cannot be read, convert it to MP3 or M4A first. A video without an audio track cannot be transcribed.

What happens while it runs

Progress moves through reading the audio, uploading, queueing, transcribing and translating. The whole transcript appears at once when it is done, with a notice; select “Open” to view it. Processing runs in the background, so you can keep using kk.

  • If processing is interrupted, the sentences already recognized are kept.
  • If the transcript file is locked by OneDrive, antivirus software or the preview pane, the last few sentences may not be saved; release the lock and import again.

Cost

  • kk Pass: deducted from your plan allowance by file duration, usually less than a live recording of the same length. If your allowance is not enough for the file, the import dialog says so.
  • Where the data goes. With kk Pass, imported audio is uploaded to the kk gateway in segments and passed to the recognition provider for transcription, then deleted as soon as it finishes, and is not written to logs or a database. With a custom API your computer sends it directly to the provider you chose. See Data and telemetry.
  • Custom API: configure a speech recognition key in Settings first, and a language-model key as well if you choose a language model for translation. Providers charge you directly.

After importing

The imported content is saved as an ordinary session marked as audio or video. You can read the transcript, generate notes, ask questions and search it just like a live recording. If you chose to translate during import, each sentence has its translation; sentences without one can be translated later from the session.

Last updated: November 15, 2026