transcribe-cli

transcribe-cli is a Rust command-line transcription pipeline built on Whisper and CTranslate2.

It supports:

• CPU-optimized transcription
• optional NVIDIA CUDA execution
• automatic Whisper model download into models/
• local media files or http/https media URLs
• streaming transcription modes
• model cleanup commands

Install

From crates.io:

cargo install transcribe-cli --locked

From a local checkout:

cargo install --path . --locked

With CUDA support:

CUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda CT2_CUDA_ARCH_LIST=Auto cargo install transcribe-cli --locked --features cuda

With CUDA + cuDNN support:

CUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda CT2_CUDA_ARCH_LIST=Auto cargo install transcribe-cli --locked --features cudnn

Usage

transcribe-cli --model small audio.mp3
transcribe-cli --model medium --stream audio.flac
transcribe-cli --model medium --live https://example.com/live.mp3
transcribe-cli --model small movie.mp4
transcribe-cli --model tiny https://example.com/audio.wav

Features

• cuda: enable CUDA support with dynamic loading
• cuda-static: enable static CUDA support
• cuda-dynamic-loading: alias for the dynamic CUDA path
• cudnn: enable cuDNN on top of CUDA

Notes

• Whisper models are downloaded automatically on first use.
• Media files are decoded through the built-in Rust pipeline; no external ffmpeg dependency is required.
• Video containers work when they include an audio track in a codec supported by symphonia.
• --live is intended for direct http/https media streams and incrementally transcribes the buffered audio as new bytes arrive.
• --live is not an HLS/DASH playlist client; it expects a direct media URL that can be read as a continuous byte stream.
• By default models are stored in models/ next to the executable unless --models-dir is set.
• Whisper decoding is handled in-project through a local wrapper around CTranslate2 sys::Whisper and Hugging Face tokenizers.
• cuda and cudnn build CTranslate2 from source through ct2rs, so the NVIDIA driver alone is not enough: a CUDA Toolkit install is required.
• cudnn also requires the cuDNN development files. ct2rs looks for cuda.h under $CUDA_TOOLKIT_ROOT_DIR/include and for cudnn.h plus libcudnn under the same CUDA root.
• ct2rs defaults to CUDA_ARCH_LIST=Common, which can include architectures removed from newer CUDA toolkits such as CUDA 13.x. This project sets CT2_CUDA_ARCH_LIST=Auto by default to avoid nvcc fatal: Unsupported gpu architecture 'compute_53'.
• If auto-detection is not what you want, override it explicitly with CUDA_ARCH_LIST=8.6 or another value supported by your GPU and CUDA toolkit.
• --locked is recommended for cargo install so published installs use the crate's resolved dependency set instead of newer patch releases.