
WhisperWeb provides a browser‑based transcription workspace that runs locally using OpenAI’s open‑source Whisper models via WebGPU or WebAssembly. Users can drop MP3, WAV, M4A, MP4, MOV or WebM files up to 200 MB, or record from a microphone, then edit the generated text and timestamps before exporting TXT, SRT or VTT. The tool also converts between SRT and VTT and extracts plain text from subtitle files, all without requiring sign‑up or sending audio to a server.