News for hackers. Stuff that matters.

A comment on OpenAI Charges by the Minute, So Make the Minutes Shorter. Parent comment

Re:OpenAI Charges by the Minute, So Make the Minutes Shorter by janalsncm on Wednesday June 25, 2025 @02:04PM (#44380162)
> I wonder if there's a way to automatically detect how "fast" a person talks in an audio file

Transcribe it locally using whisper and output tokens/sec?

[ Parent | on HN ]
Re:OpenAI Charges by the Minute, So Make the Minutes Shorter by maxall4 on Wednesday June 25, 2025 @04:19PM (#44381453)
Just count syllables per second by doing an FFT plus some basic analysis.
[ Parent | on HN ]
Re:OpenAI Charges by the Minute, So Make the Minutes Shorter by tucnak on Thursday June 26, 2025 @06:57AM (#44386128)
> FFT plus some basic analysis

Yeah, totally easier than `len(transcribe(a))/len(a)`

[ Parent | on HN ]
Re:OpenAI Charges by the Minute, So Make the Minutes Shorter by janalsncm on Thursday June 26, 2025 @07:03PM (#44392295)
Maybe not as quick to code up but way faster to calculate.

The tokens/second can be used as ground truth labels for a fft->small neural net model.

[ Parent | on HN ]