USE CASE
Transcribe audio with Whisper on a cloud GPU
Turn hours of audio into text with OpenAI Whisper large-v3 in minutes, billed per second.
Billed per second from your USD balance. Stop or terminate the machine when you are done.
Recommended GPU
RTX 4090 · RTX 3090
Whisper large-v3 needs about 10 GB of VRAM. An RTX 3090 or RTX 4090 transcribes many times faster than real time.
Template to pick
PyTorch
NVIDIA CUDA + PyTorch development environment, ready with SSH and Jupyter.
Up and running in four steps
- 1
Rent the machine
Open the catalog with RTX 4090 and the PyTorch template already selected and press Rent.
See RTX 4090 machines - 2
Connect and upload audio
Connect with the SSH command from your machine's page and copy your audio files with scp.
scp -P <PORT> audio.mp3 root@<HOST>:/root/ - 3
Install Whisper
Install ffmpeg and Whisper once.
apt-get update && apt-get install -y ffmpeg pip install -U openai-whisper - 4
Transcribe
Run Whisper on your file and download the text, then terminate the machine.
whisper /root/audio.mp3 --model large-v3 --output_format txt
Questions
Which languages does it support?
Whisper large-v3 transcribes about 100 languages and can translate into English.
Do you keep my audio?
No. Files stay on your rented machine and are deleted when you terminate it.