Sentence-aware
Sentence punctuation can close a segment rather than leaving a caption hanging across a sentence.
Turn each lesson into a caption-ready file you can review or hand off. Take the lesson's MP4, MOV, or M4V with its audio track, then queue it in AFK's Transcribe tab or run AFK's CLI on it to produce SRT or JSON.
In the app, queue the lessons in Transcribe and pick a folder with SRT as the format, so each finished caption file is saved there. From the command line, call the CLI on each lesson video with the output format you need. The command writes the transcript to stdout, so redirect it to a file beside the lesson.
$ afk transcribe lesson.mp4 --srt > lesson.srt
The command writes the transcript to stdout, so each caption file can sit beside its lesson or continue into the next production step.
AFK uses real word-level timings from the transcription result to build subtitle-sized segments. Segment boundaries follow sentence punctuation, pauses of at least 0.8 seconds, a maximum of 14 words, and a maximum duration of 6 seconds.
Sentence punctuation can close a segment rather than leaving a caption hanging across a sentence.
A pause of at least 0.8 seconds is a natural break when the words provide one.
Segments stay within 14 words and 6 seconds so a long lesson does not become one unreadable subtitle.
MP4, MOV, and M4V with an audio track. MKV, WebM, and AVI are not supported; convert those to MP4 first.
No. AFK works on files that are on your Mac. Download or export the lesson first, then drop the file into Transcribe.
Yes. SRT captions are built from the timing of each word. A caption ends at the end of a sentence, at a pause, or after 14 words or 6 seconds, whichever comes first.
Yes. AFK reads the audio in chunks of about two minutes and joins the timings into one timeline, so the captions still line up at the end of a long lesson.
No. AFK reads the audio track on your Mac and the speech model runs on your Mac. The privacy page says what does use the internet.
Use the CLI reference for flags and output streams, and the transcription guide for the app's file flow. When the library grows, the batch use case shows the shell loop. See the current AFK price.