File viewers that open the file in your browser

You have a file, you do not have the program that opens it, and the file is not yours to hand to a stranger. That is the whole problem this page solves. Each of these tools reads the bytes with the browser File API and parses them in the tab — the file is never attached to a request, so there is no queue, no size limit imposed by an upload, and no copy on anyone else's disk.

The viewers group by what kind of file you are holding. Data and databases: the Parquet viewer reads the schema and rows of an Apache Parquet file through apache-arrow; the SQLite viewer opens a .db, .sqlite or .sqlite3 file with sql.js and lets you run real SQL against it; the JSON tree viewer turns a large JSON document into a collapsible tree with copyable paths; the vCard-to-CSV converter parses a .vcf export from a phone into a spreadsheet you can actually read. Documents and slides: the PPTX viewer renders a PowerPoint deck without Office installed, and the slides viewer presents a deck full screen with keyboard navigation. Geometry and maps: the STL/OBJ viewer and the DXF viewer render 3D and CAD files with layer control, and the GeoJSON viewer plots features on an interactive map. Media metadata: the EXIF viewer shows every tag a camera wrote, the photo GPS map plots a folder of images by coordinates, the EXIF remover strips the tags again before you publish, the font previewer and glyph inspector open a TTF or OTF you have not installed, the MIDI player renders and plays a .mid file, the QR reader decodes a code from an image, and the ZIP extractor lists and unpacks an archive.

How to choose: match the file, not the label. If the extension is unfamiliar, the first question is whether the file is text or binary — JSON, GeoJSON, vCard and SVG are text and open in a tree or map viewer, while Parquet, SQLite, PPTX, STL and DXF are binary and need the specific parser. If you only need one value out of a big file, use the viewer that can filter: the SQLite viewer takes SQL, the Parquet viewer sorts and exports to CSV or JSON, the JSON tree viewer copies a path. If you need to hand the file on to somebody else, look at what the viewer exposes first — the EXIF viewer routinely turns up a home address in a holiday photo, which is exactly the case the EXIF remover exists for.

The privacy claim here is mechanical, not a policy promise. `<input type="file">` gives JavaScript a handle to bytes on your disk; reading that handle is a local operation and nothing in these pages calls `fetch()` with your file as a body. What some viewers do fetch is the parser itself — the SQLite viewer loads the sql.js WebAssembly build from a CDN the first time you use it, and the 3D viewers pull their geometry libraries the same way. That request carries the library URL and nothing else. Your file stays in the tab, and closing the tab is the whole deletion process.

Two honest limits. Everything runs in the tab's memory, so a multi-gigabyte Parquet file or a 500 MB database will exhaust the browser where a desktop tool would stream it. And a viewer is not the application that wrote the file: the PPTX viewer is for reading a deck, not for reproducing every animation and embedded object exactly.