emacs-speech-to-text-locally

emacs wrapper for the local/private speech to text nerd-dictation application

The following part of README was generated by phi4:latest LLM.

Speech-to-Text using Nerd-Dictation

This Emacs package, speech-to-text-nerd-dictation, provides a convenient interface to start the Nerd-Dictation application and insert recognized text into your current buffer. Developed by Vladimir Stavrov in 2025, this tool leverages the power of speech recognition for ease of use within Emacs.

Table of Contents

1. Installation
2. Usage
3. Customization
4. License

Installation

To install speech-to-text-nerd-dictation, follow these steps:

1. Ensure you have the Nerd-Dictation application [https://github.com/ideasman42/nerd-dictation.git] and dependent packages (including VOSK-API and models) installed.
2. Clone or download this repository to your desired location.
3. Add the following line to your Emacs configuration file (usually .emacs or init.el) to load the package:

   (add-to-list 'load-path "/path/to/speech-to-text-nerd-dictation")
   (require 'speech-to-text-nerd-dictation)

4. Optionally, bind a key to speech-to-text-nerd-dictation for quick access:

   (global-set-key (kbd "C-x C-a") 'speech-to-text-nerd-dictation)

Usage

Once installed and configured, you can activate the speech-to-text functionality using the key binding defined in your configuration (default C-x C-a if set). Here's how it works:

1. Press the hotkey to start the Nerd-Dictation listener.
2. After a beep sound, speak what you want transcribed.
3. Finish speaking with a 2-second pause, after which the recognized text will be inserted into your current buffer.

Customization

You can customize the behavior of speech-to-text-nerd-dictation by modifying certain variables in your Emacs configuration:

• Path to Nerd-Dictation Application:

  (setq s2tnd-path "~/NLP/text2speech/vosk/nerd-dictation/")

  Set the path where the nerd-dictation executable resides.

• Options for Nerd-Dictation:

  (setq s2tnd-options '(("--timeout" . "2.0") ("--output" . "STDOUT")))

  Specify additional command-line options as a list of key-value pairs, e.g., ("--timeout" . "2.0").

License

This package is distributed under the terms of the GNU General Public License (GPLv3 or later). It is free software: you can redistribute it and/or modify it under the terms of the GPL.

GNU Emacs is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

---

For further information and updates, visit the project page on GitHub: v1st-git/emacs-text-to-speech-locally.

Feel free to contribute or report issues on the GitHub repository.