Kieirra/murmure

Fully local, private and cross platform Speech-to-Text with LLM Post-processing

Murmure – Offline, privacy‑first speech‑to‑text

What it is Murmure is a desktop application that turns spoken words into written text without ever sending audio to the cloud. It runs entirely on the user’s computer and uses NVIDIA’s open‑source Parakeet TDT 0.6B v3 speech‑recognition model, which is a modern neural network optimized for fast, low‑latency transcription.

Key points

  • Local‑only processing – No internet connection is required and no telemetry is collected. Your recordings stay on your machine.
  • Multilingual – Supports 25 European languages (e.g., English, French, German, Russian, Ukrainian, etc.).
  • Open source – Released under the GNU AGPL‑v3; you can inspect, modify, and redistribute the code.
  • Cross‑platform – Official binaries are provided for Windows, macOS (Apple‑silicon and Intel), and Linux (deb, AppImage, and a quick‑install script).
  • Built with Tauri – The UI is a lightweight native window powered by the Tauri framework, which bundles the Rust‑based backend with a web‑technology front‑end.

How you use it

  1. Install the appropriate package for your OS (MSI/EXE on Windows, DMG on macOS, .deb or AppImage on Linux, or via winget).
  2. Launch Murmure; a small overlay appears.
  3. Press the push‑to‑talk shortcut, speak, and the transcription appears instantly in the overlay.
  4. The text can be copied or automatically pasted into any other application (e.g., a ChatGPT prompt, a note‑taking app, etc.).

Typical scenarios

  • Dictating code or prompts for local LLMs.
  • Hands‑free note taking.
  • Capturing ideas while working in other programs.
  • Post‑processing the transcript with a local language model for translation or grammar correction.

Installation highlights

  • Windows – MSI or setup.exe installer; also available via winget. Requires the Microsoft Visual C++ Redistributable (most systems already have it).
  • Linux – One‑line script (curl … | sh), .deb package, or AppImage. Wayland users get global shortcuts automatically on KDE/Hyprland/Sway; GNOME needs a manual shortcut.
  • macOS – Download the appropriate DMG (Apple‑silicon or Intel), drag to Applications, grant microphone & accessibility permissions, then restart.

Advanced usage

  • From version 1.8.0 the CLI can import a .murmure configuration file (murmure import config.murmure) with a --strategy merge flag for bulk deployments.
  • Future roadmap items include a richer overlay UI, CLI quit flag, draft‑mode editing, and an “LLM Connect” service to pipe transcripts directly into local language models.

Community & support

  • Issues are filed on GitHub; a contributing guide is provided.\n- Sponsorship is handled via Tipeee, and Windows code‑signing is supplied free of charge by SignPath.

Why it matters Murmure demonstrates that high‑quality speech recognition no longer has to rely on cloud APIs. By bundling a state‑of‑the‑art NVIDIA model with a privacy‑preserving desktop app, it gives users full control over their voice data while still supporting a wide range of European languages.

License: GNU AGPL‑v3 (requires derivative works to remain open source).

Related

  • Project
    theJayTea/WritingTools

    Writing Tools is a free, open‑source desktop app for Windows, Linux and macOS that lets you invoke an LLM anywhere on your computer with a hot‑key. It can proofread, rewrite, change tone, translate, summarise web pages or YouTube transcripts, and even chat with the model. You can use cloud APIs (Gemini, OpenAI, Anthropic, etc.) or run local models via Ollama, llama.cpp or Apple‑silicon MLX. The app is privacy‑focused (no telemetry, local storage of API keys) and licensed under GPL‑v3.

  • Project
    kyutai-labs/unmute

    A system that enables text LLMs to listen and speak by wrapping them in low-latency speech-to-text and text-to-speech models.

  • Project
    OpenWhispr/openwhispr

    OpenWhispr is a cross‑platform desktop app for privacy‑first voice dictation, meeting transcription, and AI‑assistant interaction. It supports offline speech‑to‑text with Whisper‑based models or cloud services, offers speaker diarisation, note‑taking with semantic search, team collaboration, and a public API for developers.

  • Project
    cjpais/Handy

    Handy is an open‑source, cross‑platform desktop app that does offline speech‑to‑text using Whisper or Parakeet models. You press a global shortcut, speak, and the transcribed text is pasted into any window. Built with Tauri (Rust + React), it runs on Windows, macOS, and Linux, supports GPU acceleration, and offers CLI flags and Raycast integration for power‑user control.

  • Project
    homelab-00/TranscriptionSuite

    A free, open-source transcription app that converts speech to text locally on your computer using various AI models for privacy and speed.