Terms of Service

The agreement

These terms are the agreement between you and Michael Lynn (“we”, “us”) covering Humanizer at rewr.it — the website, the public HTTP API, and the browser extension. Using any of them means you accept them. If you do not, do not use the service.

If you are using Humanizer for an employer or a client, you are confirming you have the authority to accept these terms on their behalf, and “you” means both of you.

You must be at least 13 years old to use Humanizer, and at least 16 if you are in the European Economic Area or the United Kingdom.

What the service does

You paste a draft. The service marks the phrases and sentence shapes that models tend to leave behind, sends your draft to a language model with instructions to rewrite around them, and returns the rewrite along with a second score and a fact audit.

The rewrite is meant to change how a draft sounds, not what it says. Every claim in the output is supposed to trace back to a sentence you wrote. That is an instruction to a model, not a guarantee — see what the rewrite is not.

Drafts and rewrites are not stored. They exist for the length of the request and are gone when it finishes. What that means in detail is in the privacy policy.

Accounts

You can use Humanizer without an account, at a lower limit. To go past that you sign in with Google, and we get your name, email address, and profile picture from Google.

You are responsible for what happens under your account, including anything done with an API key you created. If a key leaks, revoke it on your account page — that takes effect immediately. Tell us at merlynn@gmail.com if you think someone else has reached your account.

You can delete your account at any time by writing to us. Deleting it ends any active subscription and forfeits any unused credits.

Plans and limits

Every plan is capped, and the caps below are the ones the software actually enforces. A rewrite that fails or that you cancel partway still counts against your allowance, because the model has already been asked to do the work.

Monthly allowances reset at the start of each calendar month, UTC. Daily ones reset at midnight UTC. There are also per-minute burst limits and a limit on how many rewrites can run at once, which exist to keep one account from crowding out everyone else.

We may change these limits. If a change reduces what you are already paying for, we will tell you before it takes effect.

Paying

Pro

Pro is $10 per month, billed through Stripe, and renews automatically until you cancel. Cancel from the billing portal on your account page. Cancelling stops the next charge; it does not refund the current month, and you keep Pro until that month ends.

Prices are in US dollars and exclude tax, which is added at checkout where we are required to collect it. If a payment fails and stays failed, the account drops to Free.

Credits

Credits are prepaid and bought once. They do not expire. Your included monthly allowance is always spent first, and credits cover rewrites after it runs out. One credit is $0.001 of retail usage, and a rewrite is charged from the model input and output it actually measured, after it finishes.

Refunds

Credits are non-refundable once bought, and have no cash value. Monthly subscription fees are non-refundable for a period already started.

If the service billed you for something that did not work — a charge you did not make, a rewrite you were charged for that never returned — write to merlynn@gmail.com and we will sort it out. The paragraph above is about buyer’s remorse, not about our mistakes.

Where local law gives you a refund or withdrawal right we cannot contract out of, that law wins over this section.

Your drafts stay yours

You keep every right you had in the draft you paste. We claim no ownership of it, and no licence beyond the one we need to run the request: to transmit your draft to the model provider, get a rewrite back, and return it to you.

We do not use your drafts or rewrites to train models, to improve the service, or for any purpose beyond answering the request you made.

As between you and us, the rewrite is yours to use however you like. We make no claim on it and ask for no credit. Bear in mind that a model can produce similar output for different people, so nobody can promise a rewrite is unique or that it is protectable by copyright in your jurisdiction.

By submitting a draft you confirm you have the right to submit it, and that doing so does not break a confidentiality obligation you are under.

What the rewrite is not

This is the section worth reading. Humanizer is built on a language model, and language models get things wrong.

• It is not a guarantee of accuracy. The rewriter is instructed to preserve every fact and to flag any sentence it cannot preserve safely, and the fact audit exists to catch what slips. It will still sometimes miss. Read the rewrite before you send it — especially numbers, dates, names, and hedges.
• It is not a way to beat an AI detector. We do not claim the output evades detection, and anyone who does is guessing. Detectors change constantly and any text can be scored either way.
• It is not advice. Nothing the service returns is legal, medical, financial, or professional advice, whatever the draft was about.
• It is not a way to pass off work that is not yours. If the underlying work is not yours, rewriting it does not make it yours, and doing so may break your school’s or employer’s rules. That is on you.

The rewrite goes out under your name. Checking it is your job, and it is not one you can hand to the software.

What you may not do

Do not use Humanizer to:

• Produce content that is unlawful, that harasses or defames someone, that sexualises a minor, or that is designed to deceive a specific person into handing over money, credentials, or access.
• Impersonate a real person or organisation, or generate content presented as theirs without their permission.
• Submit someone else’s personal information, credentials, or confidential material that you are not entitled to share.
• Run automated traffic beyond your plan’s limits, share an API key outside your account, resell access, or work around a quota, rate limit, or paywall.
• Probe, scrape, or interfere with the service or the infrastructure it runs on, or use it to build a competing product from its outputs.
• Break the acceptable-use policy of the model provider behind the service, which applies to your drafts as well as ours.

We can suspend or close an account that does any of this, and we do not have to warn you first if the abuse is causing harm.

Availability

Humanizer is provided as it is. There is no uptime commitment, no service level agreement, and no promise that a given model, register, or feature will still be there next month. We may change, pause, or retire any part of it.

The service depends on other companies — the model provider, the database host, the payment processor, the hosting platform. When one of them is down, so are we.

If we retire the service entirely, we will give notice at least 30 days ahead where we can, and refund the unused portion of any active subscription.

No warranty

To the fullest extent the law allows, Humanizer is provided “as is” and “as available”, without warranties of any kind, express or implied — including merchantability, fitness for a particular purpose, non-infringement, and any warranty that the output will be accurate, complete, original, or suitable for what you plan to do with it.

Some jurisdictions do not allow the exclusion of implied warranties. If yours is one of them, this section applies to you only as far as that law permits.

Limits on liability

To the fullest extent the law allows, neither we nor our suppliers are liable for indirect, incidental, special, consequential, or punitive damages, or for lost profits, lost revenue, lost data, or reputational harm, arising out of your use of Humanizer — including anything that follows from sending a rewrite you did not check.

Our total liability for any claim relating to the service is capped at the greater of the amount you paid us in the twelve months before the claim arose, or $50.

Nothing here excludes liability that cannot be excluded by law, including for fraud, or for death or personal injury caused by negligence.

Indemnity

You will defend and indemnify us against claims, losses, and reasonable legal costs arising from drafts you submitted, from how you used a rewrite, or from your breach of these terms — except to the extent the claim arises from our own breach.

Ending it

You can stop using Humanizer whenever you like, cancel Pro from the billing portal, and ask us to delete your account.

We can suspend or close your account if you break these terms, if we are required to, or if keeping it open would put the service or its other users at risk. Where the reason allows it, we will tell you why and give you a chance to fix it.

Sections that are meant to outlive the agreement do: Your drafts stay yours, No warranty, Limits on liability, Indemnity, and Governing law.

Changes to these terms

We will update this document as the product changes. The date at the top says when the current version took effect, and the file lives in the repository’s history, so a change is a diff rather than a claim.

For a change that materially reduces your rights, we will give at least 30 days’ notice by email to signed-in users before it applies. Continuing to use Humanizer after a change takes effect means you accept it.

Governing law

These terms are governed by the laws of Commonwealth of Pennsylvania, U.S., without regard to conflict-of-laws rules, and the courts there have exclusive jurisdiction over any dispute — except that either of us may seek injunctive relief wherever it is needed.

If you are a consumer, this does not take away the protection of mandatory laws in the country where you live, or your right to bring a claim in your local courts.

If a provision here is unenforceable, the rest stays in force. Not enforcing something once does not waive it. These terms, together with the privacy policy, are the whole agreement between us about Humanizer.

Notices to us go to merlynn@gmail.com. Notices to you go to the email address on your account, or appear in the app if you do not have one. Our environment names the operating entity in NEXT_PUBLIC_LEGAL_ENTITY; where that is unset, this document is a draft.