scribe-cli 1.2.0

Speech-to-text CLI and system-tray app for dictating into any focused window. Local (vosk, faster-whisper) or cloud (groq, openai) backends, batch or streaming.

Scribe

Scribe is a speech-to-text CLI and tray app that pipes transcribed text into the focused window. It supports local and cloud-based APIs, batch and streaming workflows.

• Five backends, one interface. Records from your mic and transcribes via Vosk (local, streaming), Whisper (local, batch), Whisper FUTO (local, batch — ACFT-tuned for short dictations), OpenAI (cloud, batch or streaming), or Groq (cloud, batch).
• Four ways to deliver the transcript. Paste into the focused window (default), copy to the clipboard, print to the terminal, or append to a file.
• Tray or terminal. Runs as a system tray icon with a single Record button, or as an interactive terminal TUI — same menu in both.
• Hotkey-friendly. Hooks into your desktop's keyboard shortcuts via SIGUSR1 (toggle recording) and SIGUSR2 (cancel), plus built-in global hotkeys on X11 / Windows.
• Cross-platform. Tested on Ubuntu (X11 and Wayland), macOS, and Windows; works under Termux for clipboard / terminal output.

Installation

Linux / macOS:

sudo apt-get install portaudio19-dev xclip   # Ubuntu; macOS: brew install portaudio
pip install scribe-cli[all]
export GROQ_API_KEY=YOURAPIKEY                # or OPENAI_API_KEY, or skip and run local

Windows (PowerShell) — no system packages needed; sounddevice bundles PortAudio and the clipboard is native, so skip the apt/brew step:

py -m venv .venv
.\.venv\Scripts\Activate.ps1          # if blocked: Set-ExecutionPolicy -Scope CurrentUser RemoteSigned
pip install scribe-cli[whisper]        # local Whisper; or [all], or a cloud backend
$env:GROQ_API_KEY = "YOURAPIKEY"        # or OPENAI_API_KEY, or skip and run local

See the installation page for the full per-platform walkthrough (extras, Ubuntu / GNOME tray libs, the Windows quickstart).

Quickstart

scribe

This launches the system tray icon. Press Record, speak, press Stop — the transcription lands in the focused window. Scribe picks the first backend whose key / dependency is present, in order groq → openai → whisper-futo → whisper → vosk. Override the defaults or drop the tray entirely:

scribe --backend whisper --model small   # local, no API key
scribe --frontend terminal               # interactive TUI menu
scribe --mode clipboard                  # copy to clipboard, no keystroke

See the quickstart for more.

Documentation

Full documentation lives at https://perrette.github.io/scribe/:

• Installation & dependencies — PortAudio, extras, Ubuntu / GNOME tray libs, Windows.
• Quickstart — your first dictation.
• Backends in detail — model lists, streaming recipes, vocabulary biasing, the realtime model.
• Output modes & typer backends — keystroke vs clipboard, Wayland / eitype, --type-direct.
• System tray & global hotkeys — menu tree, icon states, SIGUSR1/SIGUSR2.
• Desktop entry & autostart (scribe-install) — GNOME / KDE launcher integration.
• Fine tuning & CLI reference — every scribe --help flag with examples.

From the same author

A few other open-source tools I maintain.

Scientific writing & data

• texmark — write scientific articles in Markdown and convert them to journal-ready LaTeX/PDF.
• papers — command-line BibTeX bibliography and PDF library manager.
• datamanifest — declarative, reproducible dataset management. (See also the datamanifest.toml format spec and the DataManifest.jl Julia port.)

Speech to Text (dictate) and Text to Speech (read-aloud) tools

• bard — text-to-speech reader.

Compatibility

OS	Status
Ubuntu 24.04	Primary dev platform (GNOME, X11 and Wayland).
macOS	Works.
Windows 11	Tested and working on Python 3.14 (64-bit / win_amd64). Dependencies resolve pre-built wheels; no C toolchain or Python downgrade needed.

Wayland keystroke injection is convoluted but solved.