VidverterHow to transcribe MP3 files to text
- Add your MP3 file.
- Pick the spoken language or keep automatic detection.
- Click "Transcribe", check the preview and download TXT, or SRT/VTT with timestamps.
Private by design
Most transcription services upload your recording to their servers. Vidverter runs the Whisper model inside your browser, so confidential interviews, patient notes or internal meetings never leave your computer.
Tips for the best transcript
- Choose the language if you know it; auto-detection can misjudge the first seconds.
- Use the "Accurate" model unless your device is old or slow.
- Clear audio matters more than anything: reduce background music and noise if you can.
Frequently asked questions
Are my files uploaded to a server?
No. Vidverter runs entirely in your browser and uses your device's own video encoders. Files never leave your computer or phone, so it is safe for private recordings, and there is no upload or download wait.
How accurate is it?
Vidverter uses OpenAI's Whisper speech recognition model, which is very accurate for clear speech in English and most major languages. Accuracy drops with heavy background noise, music or several people talking at once. The "Accurate" model is better than "Fast" but downloads a larger file the first time.
Is it really free?
Yes, for files up to 10 minutes, with no sign-up and no watermark. Pro transcribes recordings of any length (lectures, podcasts, meetings) and whole batches of files. On iPhone and iPad before iOS 26, and in Safari on a Mac before macOS 26, the limit is 10 minutes per file on every plan, because of browser memory limits.
Which languages are supported?
Around 100 languages, including English, Spanish, French, German, Portuguese, Italian, Dutch, Polish, Russian, Japanese, Korean, Chinese, Hindi and Arabic. Leave "Detect automatically" on, or pick the language for slightly better results.
Why does the first run take longer?
The speech model (44–80 MB on phones, up to 209 MB on computers with WebGPU) is downloaded once and then stored by your browser, so later transcriptions start right away. Processing happens on your own computer, so a newer device or a browser with WebGPU (recent Chrome or Edge) is much faster.