Speech to Text — Transcribe Audio to Text and Subtitles, Free
Transcribe recordings or your microphone into text and subtitles with an on-device Whisper model. Free, no sign-up, and your files never leave your device.
Free, with no sign-up and no watermark. The file is processed in your browser and never uploaded.
Turn recordings into text with OpenAI's open-source Whisper model running entirely on your device. Upload an audio or video file — or record from your microphone — pick the language (or let it auto-detect), and get an editable transcript plus SRT and VTT subtitle files with timestamps.
Choose Fast (≈64 MB model) or Better (≈105 MB). The model downloads once and is cached by your browser. Transcription speed depends on your computer; expect roughly real-time or faster on a modern laptop.
Step by step
- Add audio: Upload a file, or click Start recording and then Stop & use recording.
- Choose accuracy and language: Pick Fast or Better, and select the spoken language or leave Auto-detect.
- Transcribe: Click Transcribe and wait for the progress messages to finish.
- Review and edit: Correct the text in the box, or tick Show timestamps to see timed lines.
- Export: Click Copy text, .txt, .srt subtitles or .vtt subtitles.
Handy for
Meeting & interview notes
Get searchable text from recordings.
Subtitles
Create SRT captions for your videos.
Lectures
Transcribe classes for revision.
Voice notes
Turn spoken ideas into text.
Good to know
Selecting the language explicitly is faster and more accurate than auto-detect.
Clear audio with little background noise gives the best results.
Subtitle files use the original timed transcript, so export them before editing the text.
Questions
Does it work offline?
After the model has downloaded once, yes — transcription runs locally.
Is live, real-time captioning supported?
No. Record first, then transcribe the recording.
How accurate is it?
Whisper is strong on clear English and good on many other languages, but names, jargon and heavy accents may need correcting. The Better model is noticeably more accurate.
Are my audio files uploaded anywhere?
No. Decoding, editing and encoding all happen inside your browser tab. Nothing is sent to a server, and closing the tab discards everything you haven't downloaded.
Further reading
- Transcribe Audio to Text Privately, on Your Own Device — Speech to Text runs OpenAI's Whisper model in your browser: pick the language, transcribe, then export text, SRT/VTT subtitles or a PDF.
More Audio Tools tools
- Online Tone Generator — Generate sine, square, sawtooth or triangle tones, sweeps and white or pink noise, and download them.
- Online Metronome — A precise metronome with time signatures, subdivisions, accents and tap tempo.
- Audio Converter — Convert audio between MP3, WAV, M4A, OGG and FLAC, or pull the sound out of a video, right in your browser.
- WAV to MP3 — Turn large WAV recordings into compact MP3 files with the bitrate you choose.
- M4A to MP3 — Convert M4A and AAC files from iPhones, voice memos and iTunes into universally playable MP3.
- OGG to MP3 — Convert OGG, Opus and WhatsApp voice notes into MP3 files that play everywhere.