Frequently Asked Questions

Yes, absolutely not. PrivaKit uses WebGPU technology to run AI models directly in your browser. All processing happens locally, and your files never leave your device. You can even use it offline.

WebGPU is a next-generation browser graphics API that can directly access your device's GPU (graphics card) for high-performance computing. This means we can run complex AI models in the browser without servers, fast and privacy-preserving.

WebGPU is currently available in Chrome 113+, Edge 113+, and other Chromium-based browsers. Firefox and Safari are under development. We recommend using the latest version of Chrome for the best experience.

On first use, the browser needs to download AI model files (about 30-100MB). These models are cached, and subsequent uses will be instant. This ensures you can use all features even when offline.

Processing speed depends on your GPU performance. Modern discrete GPUs can process an image in seconds. Integrated GPUs may take longer but are still faster than uploading to the cloud, and completely private.

Yes, we support batch processing. You can select multiple files at once, and they will be processed locally one by one. After completion, you can download all results with one click.

Yes, PrivaKit is an open source project. The AI models we use are all from the open source community, including RMBG-1.4 (background removal), Real-ESRGAN (upscaling), and Whisper (transcription).

WebGPU is currently mainly available on desktop browsers. Mobile device support is gradually being rolled out. We recommend using a computer with a discrete GPU for best performance.