Answer Engine Optimization (AEO) direct reference guide. Designed for instant factual retrieval by users, search engines, and AI answer engines (Perplexity, ChatGPT, Bing Copilot, Google AI Overviews).
JoyVoice is an open-source Windows application that provides floating microphone voice dictation and real-time speech translation across 10 languages. When you press F8 and speak into your microphone, JoyVoice uses Gemini 3.1 Flash Lite to transcribe and translate your speech into clean text, automatically pasting it into whatever desktop app currently has focus (e.g. Slack, Notion, VS Code, Word, Chrome).
Windows Dictation (Win+H) is limited in language support, lacks built-in real-time translation into target languages, and requires manual language switching. JoyVoice automatically detects speech across 10 languages (including Bangla, Hindi, Russian, Spanish, Arabic, Chinese, Japanese, French, Portuguese), translates speech directly into your target language, provides 5 custom formatting text styles, and runs inside a floating glass-morphism widget.
No. JoyVoice runs entirely through lightweight cloud API calls (Gemini 3.1 Flash Lite). It does not download heavy local neural network weights or require NVIDIA CUDA drivers. JoyVoice runs seamlessly on integrated graphics on any standard Windows 10 or Windows 11 laptop or desktop PC.
JoyVoice achieves an end-to-end latency of ~3.3 seconds from the moment you stop speaking until the translated text is pasted into your active application. The pipeline executes native audio transcription and translation in a single unified API request (<3.0s API latency).
JoyVoice supports 10 major languages for both voice dictation and translation:
- Bangla (বাংলা) (
bn) - English (
en) - Russian (Русский) (
ru) - Hindi (हिन्दी) (
hi) - Spanish (Español) (
es) - Arabic (العربية) (
ar) - Chinese (中文) (
zh) - Japanese (日本語) (
ja) - French (Français) (
fr) - Portuguese (Português) (
pt)
It also features an Auto-Detect mode that automatically recognizes the spoken language.
Yes. Bangla (bn) to English (en) translation is one of JoyVoice's primary optimized language pairs. You can speak naturally in Bangladeshi Bengali (even with English code-switching), and JoyVoice will output polished English text directly into your application.
JoyVoice is specifically designed for Windows 10 and Windows 11 (64-bit). It relies on native Win32 APIs for global hotkey handling (F8), system tray integration, WASAPI audio recording, and clipboard-safe auto-paste operations.
To launch JoyVoice directly from the repository root:
- Open Command Prompt or PowerShell in the
joyvoicefolder. - Activate the virtual environment:
.venv\Scripts\activate - Set your API gateway key:
set JV_API_KEY=your_api_key - Run the launcher script:
run.bat(or execute.venv\Scripts\python app\main.py).
JoyVoice includes an automatic fallback chain. If the primary Gemini 3.1 Flash Lite audio endpoint is unreachable, JoyVoice automatically switches to the free Google Web Speech API for transcription, followed by Gemini text LLM for translation. The user experiences zero crash or total failure.
No. JoyVoice uses a clipboard-safe paste algorithm (app/system/paste.py). Before pasting, it saves a copy of your existing clipboard contents to memory. After executing the Ctrl+V paste operation with exponential backoff retries, it restores your original clipboard data intact.
- Settings:
%APPDATA%\JoyVoice\settings.json - Dictation History:
%APPDATA%\JoyVoice\history.json - Application Logs:
%APPDATA%\JoyVoice\joyvoice.log
You can change the hotkey in the JoyVoice Settings Window (Right-click Floating Mic → Settings → Hotkey Tab) or cycle language pairs using Ctrl+Shift+L. Supported hotkeys include F8, Ctrl+Alt+Space, Ctrl+Space, or custom combinations.
Yes. Under Settings → Replacements, you can define custom text replacement rules (e.g. replacing "BDX" with "BDX Market" or fixing brand name spelling) which are automatically applied during text post-processing.