Private by design
Add captions without uploading your video.
CaptionCandy transcribes and renders captions entirely in your browser. Your footage never touches a server — ours or anyone else’s.
Almost every caption tool works the same way: you upload your footage to their cloud, their servers process it, and you trust their retention policy. That’s a non-starter for client work under NDA, unreleased product footage, interviews, family videos — or just anything you’d rather keep to yourself.
CaptionCandy takes the opposite approach. The speech-to-text models download to your browser and run on your own hardware, and the captioned video is rendered on your device too. There is no upload step at all — which also means no waiting for a 2 GB file to crawl up your connection before work can even begin.
Zero upload, zero retention
Your video can’t leak from a server it never reached. The only data we store is your account, plan and export-minute usage.
Safe for client and NDA work
Caption confidential footage without adding a third-party processor to your compliance story.
Big files start instantly
No upload means a 2 GB 4K file starts transcribing in seconds — the wait most cloud tools call “processing” simply doesn’t exist.
Models cache after first use
The speech model downloads once and is cached by your browser, so later sessions start faster.
How it works
Open the editor and pick a file
The file is read locally by your browser — watch your network tab: there’s no upload.
Transcription runs on your device
On-device Whisper (and a dedicated Hindi engine) turn speech into timed words, right in the browser.
Render and save locally
The captioned video is composed on your device and saved straight to disk.
Common questions
How can it caption a video without uploading it?
Modern browsers can run real machine-learning models through WebAssembly and WebGPU. CaptionCandy ships the speech model to your browser and runs it there — the video file itself never leaves your machine.
Is anything sent to your servers at all?
Only account and billing data: who you are, your plan, and how many export minutes you’ve used. Footage, audio and transcripts stay on your device.
Is on-device slower than cloud tools?
It depends on your hardware — recent laptops and phones transcribe comfortably fast, and captions start appearing progressively while transcription is still running. You also skip the upload wait entirely, which for large files is often the slowest step of cloud tools.
Does it work offline?
After the app and speech model have loaded once, captioning can run without a connection. You’ll need to be online to sign in and to start an export, since export minutes are verified server-side.
Getyourvideoswatched.
Caption your next video free. Every style unlocked, no credit card, no catch.