faster-whisper
A Python implementation of Whisper transcription using CTranslate2, with CPU and GPU inference, quantization and batched transcription.
About faster-whisper
Developers need a configurable Whisper inference library for transcribing audio in their own applications.
Who it’s for
- Python developers building transcription pipelines
- Teams running Whisper models on their own CPU or NVIDIA GPU
When to consider it
Consider faster-whisper when you want to control Whisper inference, batching and quantization through a Python API.
Tradeoffs & limitations
- MIT licensed and requires Python 3.9 or newer. Current CTranslate2 GPU execution requires CUDA 12 and cuDNN 9; older NVIDIA library versions require pinning an older CTranslate2 release.
SOTA overview · Documentation-based assessment · Sources & review method
Updates
No updates shared yet.
Discussion
Newest firstAsk a question or share how you use faster-whisper.
Keep it helpful. Community rules
Loading discussion…
Sources & review method
Documentation-based assessment · Oct 30, 2026 · Prepared with AI assistance; not a hands-on benchmark.
- Official implementation, Python API and CPU/GPU requirements
- MIT licence
- Maintainer organization and owner domain
Checked by SOTA · AI-assisted documentation review. Selection advice is our assessment; verify current requirements for your deployment.
Import history & original evidence
faster-whisper official repository
A Python implementation of Whisper transcription using CTranslate2, with CPU and GPU inference, quantization and batched transcription.
Based on official pages and announcements checked on 2026-10-30. No hands-on test, performance benchmark or popularity ranking is claimed.