Privacy Policy
How this version of CaptionLeaf handles your recording and transcript.
Your file and transcript
Local files are read in your browser and are not uploaded. For an Instagram import, CaptionLeaf receives a normalized video link without tracking parameters, fetches its media from Instagram, checks it in server memory and streams it to your browser. The application does not save media files or transcripts on the server. Audio decoding, speech recognition and transcript editing happen on your device. There is no account system or cloud transcript storage. Refreshing or closing the page clears current work; download anything you want to keep.
Model downloads and browser storage
When you start transcription, the browser downloads model files from Hugging Face and its delivery services. Those services receive normal request information, such as your IP address, but not your recording or transcript. Your browser may cache the model so it can be reused. Clearing this site's browser data removes its local cached data.
Website hosting
The site is hosted on Vercel. The hosting provider may process connection and request logs. Instagram receives requests needed to fetch linked recordings. This version does not include advertising, third-party analytics or session recording.
What you control
Choose a local file or paste a Instagram link, start or cancel processing, edit the result and download it locally. CaptionLeaf does not request camera or microphone permission. Share only recordings and transcripts you intend others to see.