All products

Create ChatCut subtitles with AI

Locate a spoken line by its timestamp, correct the source wording and prepare captions once ChatCut has finished analysing the clip.

Create ChatCut subtitles with AI when an interview needs a corrected product name before the captions go to a translator. The transcription skill lets an agent identify the right asset, read its actual status and locate words at their timestamps. A spelling correction belongs in the source transcript if later captions need to inherit it.

A newly imported file may still be processing. A silent sequence may finish successfully without producing words at all. Distinguishing those cases prevents repeated submissions and keeps the editor focused on the clips that really contain speech.

What it does

  • Find a remembered phrase without replaying the entire interview.
  • Check whether the selected clip already has a usable transcript.
  • Correct source words before adjusting caption appearance.
  • Translate captions after the original language has been checked.
  • Recognise a completed no-speech result and move to the next asset.

Install and use

  1. Use an authorised project and the hosted ChatCut plugin connection. Import the asset first, with permission to send its audio for online processing.
  2. Ask, “Find the delivery-date sentence in this clip, check the product name against my spelling and then prepare captions.”
  3. Listen to a few marked passages and compare the source wording with the subtitle that viewers will see.

FAQ

Source

Made by ChatCut Inc. Original skill: transcription, published on GitHub under the Apache-2.0 licence.

View on GitHub