Voice follow: terms and privacy
Last updated: 10 October 2026
Voice follow is an optional, free tool on ShakespeareGo. You read a scene aloud and the page keeps your place. These terms explain what it downloads, what it does with your microphone, and what is and isn't kept. They add to the site's terms of service and privacy policy. By ticking the box on the voice follow page you agree to them.
What downloads, and from where
Nothing downloads until you tick the box and press Download. Then your browser fetches these files from shakespearego.com:
| File | Size | What it is |
|---|---|---|
| Speech recognition engine | 5.8 MB | vosk-browser 0.0.8, open-source software that turns speech into words inside your browser. Apache 2.0 licence. |
| English speech model | 41 MB | The Vosk small US English model (vosk-model-small-en-us-0.15) by Alpha Cephei. Apache 2.0 licence. |
| Word list | 0.1 MB | Which words in Shakespeare's plays the model can recognise. Made by ShakespeareGo. |
About 47 MB in total. On mobile data that counts towards your allowance, so use Wi-Fi if you can. Each file is checked against a fixed fingerprint (a SHA-256 hash) before it's used. A file that has been changed or damaged on the way is refused and never run.
Where it's kept, and for how long
Only in the working memory of the browser tab you're using, and only while that page is open. Nothing is installed. Nothing is written to your phone's or computer's storage: not to your files, not to your browser's storage, and not to the browser's cache of saved pages (the files are fetched with an instruction not to keep a copy). When you close the tab or leave the page, your browser throws it all away. That's why it downloads again each time you use it.
Voice follow uses no cookies and none of the browser's storage features (local storage, session storage, IndexedDB or the Cache API). It doesn't remember that you agreed to these terms, so it asks each time.
The voice follow page itself (the page, its script and its styles, about 45 KB) is sent with the same instruction not to keep a copy, so your browser doesn't save that either.
Your microphone
- Voice follow uses the microphone only after you press Start listening.
- Your browser asks your permission first. If you say no, nothing happens.
- It switches off when you press Stop listening, switch tabs or apps, lock your phone, leave the page, or after 3 minutes without hearing a line of the scene.
- While it's on, your browser shows that the microphone is in use.
- You can withdraw permission at any time in your browser's site settings for shakespearego.com. In Chrome, see Use your camera and microphone in Chrome. On an iPhone, it's in Settings, under Safari or the app you use.
Your voice
The sound from your microphone goes straight into the speech engine, which runs on your own device. It never leaves your device, it's never recorded, and it isn't sent to ShakespeareGo or anyone else.
The engine is only allowed to recognise words that appear in the scene you're reading (at most a few speeches at a time, or the whole scene while it's finding your place). Anything else, like conversation in the room, comes out as "unknown" and is discarded straight away. The few recognised words it uses to keep your place are held in memory for a few seconds and then dropped.
What we, and others, receive
Nothing about you or your voice. The voice follow page sends no information anywhere: its security settings only let it fetch the files listed above from shakespearego.com. It has no analytics, no advertising and no code from any other website.
As with any web page, your device has to ask our website for those files. The website is hosted by Cloudflare, which handles those requests, including your device's IP address, to deliver the files. See the site's privacy policy for more.
Children
Voice follow is designed for students, actors and drama groups. If you're under 13, ask a parent, carer or teacher before using it. Because nothing is recorded, sent or stored, using it doesn't share a child's voice or any personal information with us or anyone else.
It's a beta
Voice follow is free and experimental. It can lose its place, and it won't recognise every accent or unusual word. It's provided as it is, without any guarantee that it will work on your device or be available. It's your choice to use it, and you can stop at any time. Nothing in these terms affects your legal rights.
Open-source software
Voice follow uses the speech engine vosk-browser and the Vosk speech model, both released under the Apache License 2.0. Our copy of the engine is fixed at one checked version and has two changes, which stop it saving anything and stop it running code built from text. Our copy of the licence: LICENSE-vosk-browser.txt.
Changes
If we change how voice follow works in a way that affects these terms, we'll update this page and the date at the top before the change goes live.
Contact
Questions? You'll find our contact details in the privacy policy.