Stayput

136 tools in your browser. Your files are not uploaded.

Productivity

Open Source

Privacy

GitHub

Hi Product Hunt. I'm Keenan, and Stayput is the file tool I kept wishing existed. Every time I needed to turn an iPhone photo into a JPG or shrink a PDF for a form, the site I landed on wanted the file uploaded first. Then it wanted an account, or capped me at a few files a day, or watermarked the result. For a job a browser can do on its own, that never made sense. So Stayput is a static site: there is no server that receives your files. It has 136 tools for images, PDFs, video and audio, and they run in your tab: HEIC to JPG, image convert and compress, EXIF removal, background removal, PDF merge, split, compress and sign, video compress and trim, captions burned into a video, speech to text, video to MP3, and more. Drop files, get files. No account, no watermark. You don't have to take my word that your files are not uploaded. Open your browser's network tab while you use it, or load a tool, give the page a short while to save itself for offline use, then switch to airplane mode and keep going. Every tool that takes files also counts the requests this tab makes after you add them and shows you the list; requests made inside background workers, such as a model download, are not in that list. The code is MIT on GitHub. Offline: after one visit and a short wait, most tools keep working with the network off. Tools that download a model or engine, such as captions, background removal and video to MP3, need one online run first. Background removal is about 42 MB in all. Speech to text and captions download a speech model of about 77 MB (95 MB on computers with WebGPU) and the engine that runs it. Video: Most video runs on your browser's WebCodecs through Mediabunny. For old formats the browser cannot read, the page loads a decoder built from FFmpeg's libraries (LGPL) only when you convert such a file. Analytics: a self-hosted Umami counter. It sets no cookies, stores nothing in your browser and does not follow you across sites, and it records nothing if your browser sends Do Not Track or Global Privacy Control. The privacy page lists what each event contains. The host and the stats server see IP addresses; the privacy page says so too. Honest limits: video tools use your browser's own codecs, so what works depends on the browser, and re-encoding takes real time on a phone. On a phone, a video over about 200 MB (about 2 minutes of iPhone video) may run out of memory. Captions worked on my own iPhone 17 Pro, on iOS 27, in Safari; that is the one phone I can vouch for. Speech to text uses Whisper's base model: clear speech in English and other widely spoken languages comes out well, and strong accents, crosstalk and names cause mistakes. It can translate speech into English, and only into English. OCR reads English only. PDF to Word keeps the words, paragraphs and headings from the PDF's text layer, not the page layout, and there is no Word to PDF. There are no ads and no accounts. If a tool saves you time, there is a Ko-fi link on the About page; if it ever gets expensive, I would ask for donations rather than add uploads or ads. Tell me which tool is missing or what breaks for you. Code: https://github.com/keenanlk/stayput