VoxQuest is an AI-powered voice adventure that transforms text into lifelike speech. Explore different voices, experiment with speaking styles, and bring your words to life through intelligent voice technology.
VoxQuest (An adventurous journey where your words meet the magic of AI voices)
VoxQuest is a Text-to-Speech, Voice Cloning, and Emotion-based voice application.
It has a React + Tailwind frontend and a Node.js + Express backend.
First install the dependencies:
npm installFor development mode:
npm run devThe API will run on port 3001 and the frontend on port 5173.
For production mode:
npm run build
npm startThe application will run on http://localhost:3001.
Users can enter text, select a voice and speed, generate the audio, play it, and download it.
Users can upload or record a voice sample and give it a name to create a new voice.
Users can add emotion tags inside the text, such as:
[happy]
[sad]
[angry]
[calm]
[excited]
[whisper]
The application uses these tags to generate audio with different emotions.
This section contains the generated audio files and created voices. Users can delete anything they no longer need.
server/app.js
Contains the main API routes.
server/emotion.js
Handles and splits the text based on emotion tags.
server/providers/
Contains the audio providers.
mock.js
Used for generating demo audio without a real TTS API.
http.js
Used to connect the application with a third-party TTS API.
src/App.jsx
Contains the main frontend layout and tabs.
src/tabs/
Contains the separate files for each tab.
tests/
Contains the automated tests.
Copy .env.example to .env and add your API details:
TTS_PROVIDER=http
TTS_API_URL=your_api_url
TTS_API_KEY=your_api_keyIf the third-party API uses a different request format, the buildTtsBody function in server/providers/http.js needs to be updated.
If the API returns the audio as base64 inside a JSON response, configure TTS_RESPONSE_AUDIO_FIELD.
For voice cloning, configure CLONE_API_URL and CLONE_RESPONSE_VOICE_FIELD.
If no API key is provided, the application runs in Demo Mode.
In Demo Mode, it does not generate real speech. Instead, it generates synthetic tones with different pitch and speed based on the selected voice and emotion.
This mode is mainly used to test the complete application without connecting a real TTS API.
Note: History and cloned voices are stored in memory, so they will be reset whenever the server is restarted.