Convert Audio to SRT
Create timed subtitles from a recording, interview or voice track. Transcribe locally and export an editable subtitle file.
AI model will be downloaded once and cached in your browser. Download size depends on the model and backend; actual size appears during download.
Checking browser capabilities… · Checking model cache…
How it works
- Select your audio recording.
- Choose Tiny or Base and generate subtitles locally.
- Review the transcript and download SRT, VTT or TXT.
Key features
- Free audio to SRT converter.
- Automatic language detection or a selected spoken language.
- Reusable model cache stored in your browser.
Supported formats
MP3, WAV, M4A and OGG where supported by the browser audio decoder. Up to 30 minutes per file.
Private by design
Your files stay on your device. Media is processed locally and is not uploaded to our server. Whisper runs in your browser; model downloads come from model hosting or a CDN.
Frequently asked questions
Can I transcribe an MP3?
Yes, when your browser supports decoding its audio. WAV is another useful option if a codec fails.
Does this need an API key?
No. Whisper inference runs in a browser Worker and does not use a paid transcription API.
Is there a duration limit?
This MVP limits recordings to 30 minutes per file to reduce memory pressure. Device memory can impose lower practical limits.