Installation

Build dependencies

Rust (via rustup ) plus a C toolchain, cmake, pkg-config, libclang, and ALSA headers:

# Debian/Ubuntu:
sudo apt install build-essential cmake pkg-config libclang-dev libasound2-dev

# Fedora:
sudo dnf install gcc gcc-c++ cmake pkg-config clang-devel alsa-lib-devel

# Arch:
sudo pacman -S base-devel cmake pkgconf clang alsa-lib

For the authoritative list, see test-containers/Dockerfile.vulkan .

Install

cargo install voicsh

Pipe-only build (no microphone)

Skip the ALSA dependency if you only need WAV-to-text:

cargo install voicsh \
    --no-default-features --features cli,portal,model-download

GPU acceleration

By default voicsh runs on CPU. Enable GPU for ~5-10x faster transcription:

Backend	Flag	Prerequisites
NVIDIA	`--features cuda`	CUDA Toolkit ≥ 11.0
Cross-platform	`--features vulkan`	Vulkan SDK — on Ubuntu: `libvulkan-dev mesa-vulkan-drivers vulkan-tools glslc`
AMD (discrete)	`--features hipblas`	ROCm
CPU optimized	`--features openblas`	`libopenblas-dev` / `openblas`

Verify with voicsh check (shows detected GPU hardware and compiled backend).

Runtime dependencies (mic mode only)

Text injection needs one of:

Nothing extra on GNOME 45+ / KDE 6.1+ (uses xdg-desktop-portal)
wtype for wlroots compositors (Sway, Hyprland)
ydotool + ydotoold as fallback

wl-clipboard (wl-copy) is required for clipboard access.

voicsh check    # verify what's available
voicsh init     # auto-detect and configure the best backend

Pipe mode (cat file.wav | voicsh) has no runtime dependencies.

Architecture