Speech Data Builder — TTS & STT Speech Dataset Creation Tool

Speech Data Builder is a free, open-source, browser-based tool for creating high-quality speech datasets for Text-to-Speech (TTS) and Speech-to-Text (STT) AI model training. Upload audio, transcribe it with AI or by hand, normalize the text, and export to LJSpeech, Common Voice, CSV, or JSON — all without leaving your browser.

Key features

• AI transcription with Google Gemini and OpenAI Whisper
• Waveform editing with region selection, trim, and playback
• Automatic text normalization for TTS training
• Import LJSpeech, CSV, TSV, JSON, and TXT transcripts
• Export to LJSpeech, Common Voice, CSV, and JSON
• Batch processing and installable offline PWA