faster-whisper
A faster version of Whisper built on CTranslate2.
Open the source (opens github.com) https://github.com/systran/faster-whisper
Before you use it
- Running it on a GPU needs matching CUDA libraries; the README lists the versions.
What it is
Models from SYSTRAN. Good for:
- Transcribe a lot of audio on a budget
Level: builder
How to get it
Install
pip install faster-whisper
Copied from the official README or docs
Related
-
faster-whisper-large-v3
Modelsautomatic speech recognition model by Systran · ctranslate2
Systran/faster-whisper-large-v3 -
Whisper
ModelsOpenAI's open speech-to-text model.
pip install -U openai-whisper -
automatic speech recognition model by nvidia · nemo
nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B -
cohere-transcribe-03-2026
Modelsautomatic speech recognition model by CohereLabs · transformers
CohereLabs/cohere-transcribe-03-2026 -
automatic speech recognition model by nvidia · nemo
nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B-NVFP4 -
Kimi-Audio-7B-Instruct
Modelstext to speech model by moonshotai · kimi-audio
moonshotai/Kimi-Audio-7B-Instruct