Your files stay on your device.

Video, audio and subtitle files selected in these tools are read and processed locally in your browser. The tools do not upload your media or subtitle contents to our server.

Local speech recognition

Whisper inference runs in a browser Web Worker. Before the first transcription, model assets are downloaded from Hugging Face model hosting and its delivery network. These requests include normal network information, such as your IP address, but do not contain your selected media.

Browser storage

Downloaded AI model assets are saved using Cache Storage when available. Use Clear cached AI model in a transcription tool to remove that cache. Your browser may clear cached data under storage pressure. Closing or reloading a page clears its current editing state; save your work by downloading it.

Hosting and analytics

Normal website and model requests may be logged by the hosting providers. This MVP includes no analytics, account system, cloud sync or advertising trackers. If analytics are added later, they must not receive media, subtitle text or file contents.

Practical limits

Local processing still depends on your browser, device memory and codec support. This privacy description concerns the behavior of this application; it does not describe browser extensions or other software installed on your device.