whis 0.5.9

Voice-to-text CLI using OpenAI Whisper API
whis-0.5.9 is not a library.

Why?

  • Built for AI workflows — speak your prompt, paste to Claude/Copilot
  • Cheap — ~$0.006/minute via OpenAI Whisper or Mistral Voxtral (no local GPU)
  • Simple — record → transcribe → clipboard
  • Multi-provider — choose between OpenAI Whisper or Mistral Voxtral

Quick Start

cargo install whis
whis config --openai-api-key sk-your-key-here  # or --mistral-api-key
whis

Usage

One-shot mode:

whis    # Recording starts, press Enter to stop

Hotkey mode (background service):

whis listen                    # Global Ctrl+Shift+R anywhere
whis listen -k "super+space"   # Custom hotkey
whis status                    # Check if running
whis stop                      # Stop service

Configuration:

whis config --openai-api-key sk-...   # Save OpenAI API key
whis config --mistral-api-key ...     # Save Mistral API key
whis config --provider mistral        # Switch to Mistral Voxtral
whis config --language en             # Set language hint (ISO-639-1)
whis config --show                    # View current settings

Requirements

  • API key from OpenAI or Mistral
  • FFmpeg (sudo apt install ffmpeg or brew install ffmpeg)
  • Linux (X11/Wayland) or macOS

For hotkey mode (one-time setup on Linux):

sudo usermod -aG input $USER
echo 'KERNEL=="uinput", GROUP="input", MODE="0660"' | sudo tee /etc/udev/rules.d/99-uinput.rules
sudo udevadm control --reload-rules && sudo udevadm trigger
# Logout and login again

Prefer a GUI?

See whis-desktop — same functionality, with system tray.

License

MIT