File Drop
Send files of any size, browser to browser. No server, no upload.
A genuinely serverless peer-to-peer transfer. Files stream directly between two browsers over an encrypted WebRTC channel — there is no signaling server, no cloud and no account.
- No size limit — files stream, never stored
- End-to-end encrypted WebRTC DataChannel
- Invite codes replace the signaling server
Features
What File Drop does
Truly serverless
Normal P2P apps still use a signaling server to introduce the peers. File Drop replaces it with a code you share over any channel you like.
Any size, any type
Multiple files per session, drag & drop, no size limit — files are chunked at 64 KB and streamed with backpressure handling.
Encrypted by default
The connection is a WebRTC DataChannel secured with DTLS. Nothing passes through a TaskGo server because there isn't one.
Live progress on both sides
Per-file progress bars, a connection status pill (waiting / connecting / connected / done) and auto-save on arrival.
Friendly codes
Invite and reply codes are validated by prefix (TASKGO-INVITE / TASKGO-REPLY) with clear error messages if something's pasted wrong.
Under the hood
How it works
No magic, no server. Here's what actually happens when you use it.
- 01
Create an invite code
The sender picks files and gets a code — the WebRTC offer (SDP + ICE candidates), base64-encoded.
- 02
Share it anywhere
WhatsApp, email, Slack — any channel works. The receiver pastes it and gets a reply code back the same way.
- 03
Connect directly
The sender pastes the reply, the browsers connect peer-to-peer, and the files stream across.
Transparency
Every permission, explained
Chrome shows you a list at install time. This is the list, and why each item is there.
| Permission | Why it's needed |
|---|---|
| storage | Cache the announcement feed for the What's new list. |
| alarms, notifications | Check the TaskGo announcement feed every 6 hours. |
Full details in theprivacy policy.
Honest notes
Known limitations
Things it can't do, so you don't have to find out the hard way.
- Google's public STUN server is contacted once so each browser can learn its own public address — it never sees any file data.
- There is no TURN relay: if both peers are behind symmetric NATs the connection may fail. That's the honest trade-off for having no server at all.
More from TaskGo
