# Transcribe

default cuda/torch ~43% utilization
1080* i58600k 16gb ram
60min pod cast

**Medium model**
TotalMinutes: 19.5210157

**Small model**
TotalMinutes: 10.11020157

### Datacrunch.io

**Medium model**
TotalMinutes: approx 10~

## Costs

https://www.reddit.com/r/MachineLearning/comments/xl7mfy/d_some_openai_whisper_benchmarks_for_runtime_and/


## 30 sec file

## Random

https://deepgram.com/pricing/ <- 0.78$ uz one hour transcribing lmao
https://huggingface.co/openai/whisper-medium