xifan2333/fcitx5-vinput

Voice input for Fcitx5 — local and cloud ASR, LLM rewriting, cross-distro packages

What it solves

It provides voice input capabilities for the Fcitx5 input method framework on Linux, allowing users to enter text via speech instead of typing. It addresses the need for a flexible voice-to-text system that can work either offline for privacy and speed or via cloud services for higher accuracy.

How it works

The project integrates Automatic Speech Recognition (ASR) into Fcitx5. It supports local ASR using sherpa-onnx models or cloud-based ASR through providers like OpenAI, ElevenLabs, and others. Beyond simple transcription, it can use Large Language Models (LLMs) to post-process the transcribed text for error correction, formatting, or translation. It also features a "Command mode" where users can select text and speak instructions to modify it.

Who it’s for

Linux users who use the Fcitx5 input method and want to integrate voice typing and voice-driven text editing into their workflow.

Highlights

  • Hybrid ASR: Supports both offline local models and various cloud ASR providers.
  • LLM Integration: Uses LLMs for post-processing tasks like formatting and translation.
  • Flexible Triggers: Offers both toggle-based recording and push-to-talk modes.
  • Command Mode: Allows users to modify selected text using voice instructions.
  • Broad Distribution: Provides packages for Arch, Fedora, Ubuntu/Debian, Nix, and Flatpak.

Related

  • Project
    peteonrails/voxtype

    Voxtype is an open‑source, MIT‑licensed Linux app that provides local, real‑time voice‑to‑text dictation. It runs fast CPU models (Cohere Transcribe 9‑11× realtime) and supports GPU/Intel NPU acceleration, nine interchangeable transcription engines, multilingual support, meeting‑mode transcription, and deep desktop integration (Wayland/X11, hotkey binding, auto‑pause media, waveform OSD). All processing stays on the machine unless the user opts for a remote server.

  • Project
    BryceWG/BiBi-Keyboard

    An AI-powered voice input method for Android that supports multiple cloud and local ASR engines with LLM-based text post-processing.

  • Project
    goodroot/hyprwhspr

    A native Linux speech-to-text tool providing fast, private, and system-wide dictation using local or cloud-based AI models.

  • Project
    VocaHQ/vocalinux

    A privacy-focused voice dictation app for Linux that converts speech to typed text locally on-device using engines like Whisper and VOSK.

  • Project
    akshayaggarwal99/jarvis-ai-assistant

    An open-source voice dictation tool that converts speech to text anywhere on your system, supporting local transcription and LLM post-processing for privacy and speed.