Transcribe audio or video to text
Turn speech into text you can copy, plus subtitles. English, Danish and many other languages.
Your recording stays on your device. Only the speech model is downloaded.
How it works
- Choose a recording or a video. Common formats like MP3, M4A, WAV, MP4 and MOV all work.
- Pick the language that is spoken, so the tool knows what to listen for.
- Press Transcribe. When it's done, copy the text or download it as a text file or SRT subtitles.
Good to know
The transcription runs on your own device with Whisper, a speech-recognition model made by OpenAI. The first time you use it, your browser downloads the model, not your files, once. After that it starts from your browser's cache.
How fast it goes depends on your device. It is quickest on a computer with a recent graphics chip. Without one, expect the transcription to take about as long as the recording, or longer on older machines and phones. Keep the tab open while it works.
It's a good first draft, not a certified transcript. Names and unclear speech can come out wrong, and it does not tell speakers apart. Read it through if accuracy matters.
Questions
Is my recording uploaded?
No. The audio stays in your browser on your own device. The only download is the speech model, the first time you use it.
Does it work for Danish?
Yes. Choose Dansk as the language. For Danish, use Most accurate if your computer offers it, or Standard otherwise. The Quick model makes many more mistakes in Danish.
How long can the recording be?
Long recordings take proportionally longer and need more memory. In our test a 15-minute recording took about 5 minutes on a laptop with the Quick setting. An hour should be fine on a computer with plenty of memory. On a phone, stay under about 20 minutes.
Can I make subtitles?
Yes. Download the SRT file, which most video editors and players can read. The tool times each line for you.
Is it free?
Yes. No account, no minute limit and no watermark.