Privacy Policy

The short version

Squint sends the region you capture to OpenAI to be read, and shows you the answer. We do not keep your captures after the answer has been returned, we do not use them to train models, and your conversation history stays on your own machine.

The application carries no analytics, no telemetry and no crash reporting. This website sets no cookies and runs no analytics either. We do not sell anything about you to anybody, and we never will.

Who is responsible for your data

HEYSQUINT LLC, a Michigan limited liability company, registered at 1312 Armstrong Dr, Chelsea, MI 48118, is the controller of the personal data described here. For anything about this policy, or to exercise any of the rights below, email [email protected].

Squint Plus is sold in the United States only, but free accounts are not restricted by country. So this policy is written to work for a reader in the UK or the EU as well, and the rights below are theirs whether or not they have ever paid us anything.

What we hold

Your email address, in two forms: as you typed it, so that mail reaches you the way you asked, and normalised, so that one mailbox cannot quietly become several free accounts.

Whether that address has been verified, and — while a sign-in link is outstanding — a hash of that link and the time it expires. The link itself is never stored.

If you use Continue with Google, the account identifier Google gives us for you. We ask Google for your email address and nothing else: not your name, not your picture, and no access to anything in your Google account.

A device identifier generated by your installation: a random value made on your machine, not derived from your hardware. It does two jobs. It limits how many free accounts one machine can create, and it is how a computer signing in for the second time keeps the place it already had instead of taking another.

The computers you are signed in on. For each one: that identifier, when it was signed in, when the credential on it was last used, and what the machine calls itself. The name is the computer's own name as your operating system reports it, so if you have named your machine after yourself, that is the name we hold. It is shown on the device list in Squint and on heysquint.com, both of which need your credential to read, and it is in the copy of your data you can download.

The identifier of the computer that asked for the most recent sign-in link, so that opening that link on a phone signs in the computer that asked for it rather than the phone.

Your plan, and — if you subscribe — a Stripe customer and subscription reference, the status of that subscription, and the dates of your billing period. Stripe holds the payment details, not us. If Plus was given to you with an activation key rather than bought, the date it runs until and the record that your account redeemed that key.

One record per turn: when it happened, which model tier it used, what it cost, and whether it was a follow-up. Never the image, and never the answer.

A record of the credit allowance each payment or activation key bought and which invoice it came from, so that a billing question has an answer.

The access tokens your installations hold, stored only as hashes, with a label, which computer each belongs to, when it was created, when it was last used, what it is allowed to do, and when it expires. A stolen copy of our database is a list of hashes rather than a set of working credentials.

Your captures

A capture is uploaded to our proxy, passed to OpenAI, and discarded once the answer has been sent. It is not written to our database and not retained. If a conversation continues, your own installation re-sends the image — the server keeps nothing between requests, which is what makes the sentence above true rather than aspirational.

Alongside the image we tell OpenAI which family of operating system the capture was taken on — Windows, macOS or Linux, and nothing more precise. It is there so that an answer about a settings panel or an error names the right steps for the system you are actually on, and it is deliberately coarse: no version, no build number, and never your computer's name.

OpenAI processes captures as our processor under their API terms, which state that data sent to the API is not used to train or improve their models.

OpenAI does generate abuse-monitoring logs for API usage and retains them for up to 30 days. That is a control on their side and we do not switch it off, so the sentence above is about our systems: for up to thirty days after a capture, a copy of it exists in OpenAI's logs, where it is used to detect misuse of their API and for nothing else.

On Linux the desktop's screenshot portal writes a file to disk before Squint can read it. Squint reads that exact file once and deletes it immediately.

When the application talks to us

Three things, and nothing else. A capture, when you take one. Your account — signing in, and reading or changing it from the Account pane, which is where your usage, your plan and your devices come from. And a check for a new version each time Squint starts.

The version check asks our server for one small file. It happens whether or not you have an account — including if you chose Not now and never made one — and, like any web request, it carries your IP address and tells us which version and platform is asking. Nothing else rides along with it, and nothing is stored against you.

There is no analytics, no telemetry and no crash reporting anywhere in Squint. We do not know how often you capture, what your hotkey is, which theme you chose or which features you use, and the reason is not policy but plumbing: nothing in the application sends us any of it.

What is kept in your browser

heysquint.com stores a small number of values in your browser's local storage and sets no cookies: the address you signed in with, an address awaiting a link, the scoped token that lets this browser read your own account, a random identifier for this browser, and your light or dark theme preference.

None of it is sent anywhere except to us, and clearing your browser's storage for this site removes all of it. Signing out removes the first three.

Servers, logs and IP addresses

Our servers see your IP address, as any server does. It is used to rate-limit the handful of endpoints that need no account — signing up, signing in, and requesting a link — so that one caller cannot mail-bomb somebody else's inbox. Those counters are held in memory, expire within the hour, and are never written to our database.

Ordinary web server logs — the request, the response, and the address it came from — are kept for up to 30 days and then discarded. They are used for security and diagnosis, are not joined to your account, and are not used to build any profile of you.

Why we are allowed to hold it

If you are in the UK or EU, we rely on three lawful bases. To provide the service you asked for and to take payment for it, we are performing our contract with you. To keep the service working, secure and un-abused — the usage limits, the device rule, the rate limits — we have a legitimate interest, and we have weighed it against the very small amount of data each of those rules needs. To keep records of what we sold and to whom, we are meeting a legal obligation.

We do not rely on consent for any of this, which is also why there is no cookie banner: there is nothing here to consent to.

How long we keep it

Account data for as long as your account exists. Usage records are kept while the account exists, because they are what enforce the limits and what show you your own history.

When you delete your account, all of it is removed. One exception, kept deliberately: the record that a device created an account in a given month survives, with the account reference removed, so that deleting and recreating cannot be used as an unlimited supply of free accounts. What is retained is a random identifier your own installation generated and a month, and it cannot be linked back to you.

Billing records are kept for as long as tax and accounting law requires us to keep them, which is longer than the account itself.

Who else processes it

OpenAI, to generate answers. Stripe, to take payment. Resend, to send sign-in email. Cloudflare, which sits in front of our servers and terminates the connection your browser makes. Oracle Cloud, which hosts them. Google, only if you choose Continue with Google.

Each of those is a processor acting on our instructions, or — in Stripe's and Google's case — a service you are also dealing with directly. Nobody else receives your data, nothing about you is sold, and none of it is used for advertising.

Where your data goes

We are established in the United States and our servers and processors are there. If you are in the UK or EU, using Squint means your personal data is transferred to and processed in the United States.

Where that transfer needs a safeguard, we rely on the standard contractual clauses and the UK addendum in our agreements with the processors above, and on the EU–US and UK–US Data Privacy Framework where a processor is certified under it.

Your rights

You can download everything we hold about you, and delete your account and its data, from Account in Squint's own window. Both take effect immediately, and neither requires you to ask us.

If you are in the UK or EU you also have the right to be told what we hold and why, to have it corrected, to object to processing carried out on the basis of our legitimate interests, to ask us to restrict it, and to receive it in a portable form. Exercising any of them costs nothing, and we will answer within one month.

If you are in California you have the right to know what is collected and why, to have it deleted or corrected, and to opt out of the sale or sharing of your personal information. We do not sell or share it, so there is nothing to opt out of. We will not treat you differently for exercising any of these rights.

If you think we have got something wrong, tell us first. You can also complain to a supervisory authority — the Information Commissioner's Office in the UK, or the data protection authority of the EU country you live in.

Children

Squint is not for anybody under 18 and we do not knowingly collect data from children. If you believe a child has created an account, email us and we will delete it.

How it is kept safe

Everything travels over TLS. Access tokens and sign-in links are stored as hashes rather than in the clear. Card details never reach us at all. Access to the database is limited to what running the service requires.

No service can promise it will never be breached. If one happens and it puts you at risk, we will tell you and the relevant authority, as the law requires.

Changes to this policy

If we change what we collect or what we do with it, we will update this page and, where the change is significant, tell you by email before it takes effect. The date at the top is when the current version took effect.

Contact

Email [email protected], or write to HEYSQUINT LLC, 1312 Armstrong Dr, Chelsea, MI 48118.