Local Drive has a web client. It reaches the same server as the apps, over the same API, and holds nothing of its own. Point a browser at it, give it the address of your server, and sign in.

It exists for the cases where installing something is the wrong answer: a borrowed laptop, a locked down work machine, a phone that is not yours, or simply wanting to check one file without putting an app on the device.

What it is

A separate client, not the Flutter app compiled for the web. It is a Next.js application in web-client/, and it is a full port: the same screens, the same two languages, the same design, the same behaviour.

The server does not know or care which client is talking to it. Anything you do in a browser shows up in the apps, and the other way round, because both are views onto one server.

What works

Everything the apps do, with three exceptions listed below.

Signing in, including the second factor and waiting for device approval.
Browsing, searching, opening folders, and the trash.
Uploading, including several files at once, with progress.
Downloading, previews for images, video, audio, PDF, text, code and markdown, and the gallery grouped by month.
Sharing, both a link anyone can open and giving access to somebody who has an account on the server.
Renaming, starring, moving to the trash, restoring, and version history.
Keeping a file for offline use.
Every settings screen: account, devices, language, two factor, downloads, and, for an admin, people and server settings.
On a desktop: rubber band selection, shift and control clicking, dragging files onto a folder, dragging files in from the desktop, and the keyboard.

What it cannot do

Three things, and all three are the browser's limits rather than decisions.

Uploads are held in memory. A browser cannot write to a scratch file, so an upload interrupted by closing the tab starts again. The apps resume after a restart. For anything large, use an app.

Nearby sharing does not work. It needs to see the local network, which a page cannot do.

Offline is a convenience, not a copy. Files marked for offline are kept in the browser's cache storage. They survive a reload and a closed tab, and they open with the server switched off, but a browser may evict them under storage pressure. Anything that must survive belongs on the server.

Where the files go

Nowhere. The client holds no data. Everything is on your server, and clearing the browser's storage signs you out and removes the offline copies, nothing more.

Running it yourself

The web client is a static site. Build it and host it anywhere, including on the same machine as the server:

cd web-client
npm install
npm run build
npm start

It asks for a server address on first run, so one build works for any server.

Two things to know about hosting it elsewhere:

The server answers any origin by default, so a client hosted on another domain works with no configuration. Setting CORS_ALLOWED_ORIGINS narrows that to a list you choose.
A page served over HTTPS cannot talk to a server over plain HTTP. That is a browser rule rather than one of ours. If the client is on HTTPS, give the server a domain so it has HTTPS too. See HTTPS.