VoiceSketch is a generative AI prototyping tool that instantly converts your spoken UI ideas into production-ready HTML + Tailwind CSS components.
Skip the wireframing.
Skip the CSS debugging.
Just speak and watch the UI appear.
- 🎙️ Voice-to-Code Engine — Converts natural speech into structured UI descriptions.
- 🧠 Powered by Gemini 2.5 Flash / 2.0 Flash — Generates clean, responsive HTML + Tailwind CSS.
- 📱 Live Interactive Preview — Renders components instantly inside Streamlit.
- 🎨 Tailwind CSS Native — Utility-first responsive design by default.
- 🛡️ Robust Architecture — Automatic text fallback, Apple Silicon audio fixes.
- 📥 One-Click Export — Download HTML output immediately.
- Frontend: Streamlit
- AI Model: Google Gemini API
- Audio: SpeechRecognition, PyAudio
- Styling: Tailwind CSS injected into output
- Backend: Python (
generator.py)
git clone https://github.com/yourusername/voicesketch.git
cd voicesketch