Privacy
Last updated 9 September 2026
This covers both this website and the LoudGPT app. The short version: we keep as little as we can, we do not sell anything to anyone, and the app is built so that most of what you type never reaches us at all.
The website
Your email address, if you type it into the waitlist form. That is the only thing this site asks you for. Alongside it we store the date you signed up and which of the two forms you used, so we can tell what is working.
We do not store your IP address. To stop one machine flooding the form, we keep a counter against a one-way hash of the address, salted with a secret. The address itself is never written down and the hash cannot be turned back into one.
The site sets no cookies, and runs no analytics, tracking pixels, session recording or advertising. The only third-party requests a page makes are for the web fonts it uses.
We use your address to email you when a place opens up, and for nothing else. We do not sell, rent or share the list.
The app: your questions
With your own API key, your questions never touch our servers. This is the normal way to use LoudGPT. The app talks to the AI provider you chose — Google, Anthropic or OpenAI — directly from your device, using your key. We cannot see those questions or answers, because they never come near us. What the provider does with them is covered by that provider’s own privacy terms, not ours.
If you sign in and use our key instead, your question passes through our server on its way to the provider. That is the trade for not needing a key of your own. We do not store it and we do not log it: the request is forwarded, the answer is returned, and nothing is written down. We count that you asked — as a number against a salted hash of your account id — so the limits can be enforced.
The app: reading answers aloud
Speech is the one part we always pay for and therefore the one part that always passes through us. To read an answer aloud, the app sends that answer’s text to our server, which turns it into audio using Google Cloud Text-to-Speech. Nothing is stored. The audio goes straight back to your device and is kept there so replaying costs nothing. As with the website, requests are counted against a salted hash and never against an address we keep.
The app: what stays on your device
- Your API keys, in the operating system’s secure store — the iOS Keychain or the Android Keystore. They are not synced anywhere and we never receive them.
- Your chats, playlists and settings. There is no account-backed history and nothing is uploaded.
- The audio we render for you. Cached on the device so replaying an answer needs no second request.
Uninstalling the app removes all of it. We have no copy to delete on your behalf.
The app: crash reports
Release builds send crash reports to Sentry so we can fix what breaks. These are deliberately narrowed: console output is dropped entirely, and query strings are stripped from any URL in a report. Both are there because this app handles question text and API keys while being debugged, and a crash report is not somewhere either belongs. Development builds report nothing, and performance tracing is off.
Where the website’s data is kept
The waitlist lives in a Postgres database hosted by Neon, on servers in Singapore. The site and the speech endpoint run on Vercel. Speech is generated by Google Cloud, and crash reports go to Sentry. Each handles data on our behalf under its own terms.
How long we keep things
The waitlist, until you ask us to remove you or until it has served its purpose and we delete it. Rate-limit counters, which contain no identity, for a matter of days. Crash reports, for Sentry’s retention period. Questions and answers are not kept at all.
Having your data removed
Email support@contactout.com and we will delete your address from the waitlist. You do not have to give a reason, and you can also ask what we hold about you. Every email we send has an unsubscribe link. For anything the app stored, uninstalling it is the deletion — none of it was ever anywhere else.
Children
LoudGPT is not aimed at children, and we do not knowingly collect anything from anyone under 13.
Changes
If what we collect changes, this page changes with it and the date at the top moves. We will not quietly start collecting something this page says we do not.
Contact
Questions about any of this go to support@contactout.com.