Free speech to text, right in your browser
Turn your voice, voice notes and recordings into text with the open-source Whisper model running on your own device. No sign-up, no uploads, no limits.
This is a quick way to try speech to text. Your audio never leaves your browser, and the Whisper model is downloaded once and cached. A browser can't use your computer's full power, so for faster transcription, smart formatting and typing straight into any app, get the free OiPer desktop app.
How to transcribe audio to text
Voice to text in three steps, for a quick voice memo, an interview or a lecture recording.
1. Record or upload
Press "Record" and speak, then press "Done". Or upload an audio or video file: MP3, WAV, M4A, OGG, WebM, FLAC or MP4 all work in modern browsers.
2. Let Whisper transcribe it
The first time, your browser downloads the open-source Whisper model and caches it, so later visits start right away. Whisper detects the language on its own and works through long recordings in 30-second pieces.
3. Copy or download the text
Fix anything in the text box if you need to, then copy the transcript or save it as a .txt file.
Want this in every app?
This page is great for a single recording. If you’d rather talk than type all day, OiPer is a free desktop app that turns speech into text wherever you’re writing.
Types into any app
Hold your hotkey, speak, and release. The text appears where your cursor is: email, docs, chat, browsers and code editors.
Free and unlimited on your computer
Local Whisper transcription costs nothing and needs no account. Audio, transcripts and history stay on your device.
Bigger models, faster on your GPU
Pick larger Whisper models for better accuracy. OiPer uses your GPU when you have one and runs on Windows, macOS and Linux.
Questions, answered
What people ask before they transcribe audio in the browser.
Is this speech to text tool really free?
Yes. There is no account, no watermark and no limit on how often you use it. Transcription runs on your own device, so there is no server cost to pass on to you.
Is my audio uploaded anywhere?
No. Your recording or file is processed inside your browser and never leaves your device. The only download is the Whisper model itself, fetched once from Hugging Face and cached by your browser.
Which audio and video formats can I transcribe?
Anything your browser can play: MP3, WAV, M4A, AAC, OGG, WebM and FLAC, plus the audio track of most MP4 and MOV videos. If a file will not open, convert it to MP3 or WAV and try again.
Which languages does it support?
Whisper understands around 99 languages and detects the spoken language automatically. Accuracy is best for English and other widely spoken languages, and larger models are more accurate: Whisper Large v3 Turbo is the most accurate, Whisper Tiny the fastest.
How long can a recording be?
There is no fixed limit. Long audio is transcribed in 30-second pieces. Speed depends on your device: browsers with WebGPU run much faster, while on CPU an hour of audio can take a long time.
How is this different from the OiPer desktop app?
This page transcribes one recording or file at a time. The OiPer app works system-wide: hold a hotkey, speak, and the text is typed into whatever app you are using on Windows, macOS or Linux, with larger Whisper models and GPU acceleration, free and unlimited.