Small tool. Open foundations.

Your files stay yours

Photos and PDFs are decoded and processed in your browser. We do not upload, store, or collect them. Closing this page releases the file session. The site serves its own AI models and runtime files; ordinary server and Cloudflare request logs may include connection metadata such as your IP address, but never your file contents. There are no analytics scripts or accounts in this application.

Free to use

All processing and full-resolution exports are free, with no credits, watermarks, subscriptions, or paid inference APIs. The site operator provides the hosting and domain. “Full resolution” preserves the uploaded image dimensions; it does not increase the model’s segmentation resolution or invent image detail. Complex hair, glass, and low-contrast edges may need further editing.

Credits

• U²-Net by Xuebin Qin and contributors — Apache 2.0; U2NET and U2NETP models.
• rembg — MIT; model distribution and preprocessing reference.
• ONNX Runtime — MIT; in-browser model inference.
• Next.js, React, Tailwind CSS, and Lucide icons — MIT / ISC.
• PDF-Lib — MIT; PDF creation and editing. Mozilla PDF.js — Apache 2.0; local PDF rendering.
• Example dog photograph — Unsplash. Original photo served locally.

Full notices are included in the source repository’s THIRD_PARTY_NOTICES.md and licenses directory.