Fact · Speech recognition
13minutes
The faster-whisper benchmark measured how long each implementation took to transcribe a 13-minute recording.
Checked against its source on October 17, 2026
In the source's own words
“to transcribe [**13 minutes**](https://www.youtube.com/watch?v=0u7tTptBo9I) of audio”
faster-whisper: README (GitHub, SYSTRAN), SYSTRAN, GitHub (SYSTRAN/faster-whisper), 2026, Benchmark.
Open the original- What it measures
- The README compares openai/whisper at v20240930, whisper.cpp at v1.7.2, transformers at v4.46.3 and faster-whisper at v1.1.0.
- What it does not show
- One recording, one set of software versions. The README's own tables are the only evidence here, so treat them as a snapshot.