Fact · Transcription
The openai-whisper command-line tool has a --word_timestamps option, which the source marks as experimental. It extracts word-level timestamps and refines the results based on them.
Checked against its source on October 17, 2026
In the source's own words
“(experimental) extract word-level timestamps and refine the results based on them”
whisper/transcribe.py: decoding defaults and command-line options (openai/whisper), OpenAI, GitHub (openai/whisper), 2026, Command-line argument parser, --word_timestamps option.
Open the original- What it measures
- This is the help text for the option in the file that the whisper command runs. The option defaults to False, so word timestamps are off unless you switch them on. The line-length options described in the next facts need this option turned on.
- What it does not show
- The source labels the option experimental, so check the timing of its output against your own audio before you publish the captions.