Whisper Web is a browser-based transcription tool powered by OpenAI’s Whisper model, running locally in the browser with WebGPU acceleration. Open the page and transcribe immediately with no installation, account, or API key: use microphone input, file upload, or browser audio, with automatic language detection across ninety-nine languages. The free plan processes audio entirely on the device, so nothing leaves the computer, and covers files up to 200 MB or twenty minutes with no limit on the number of transcriptions. Transcripts export as TXT, SRT, VTT, or JSON, and once the model is downloaded on the first visit, the free plan works completely offline. The Unlimited plan moves processing to the cloud for files up to ten hours or 5 GB, supports batch uploads of fifty files at once, and adds cross-device sync and priority processing, at $20 a month or $10 a month billed yearly. It suits students, journalists, podcasters, and creators turning recordings into searchable, editable text without uploading sensitive audio.