Our Story

About UnAI My Text

We built this because every AI humanizer we tried was either locked behind a paywall, inaccurate, or turned clear writing into mush.

We wanted one that actually works — free, with no tricks, no credit card prompts, and no output that needs editing afterward. So we made it.

What we stand for

Our Mission

Three principles that guide every decision we make — from how we build the model to how we handle your text.

Accessible

Free humanization for everyone. No credit card required, no trial period, no “upgrade to unlock.” The full tool is available from the first second you land on the page.

Accurate

We test against GPTZero, Turnitin, and Copyleaks every month. If pass rates drop, we retrain before the next update ships. You see exactly what we measure.

Private

Zero data storage, zero logging. Your text is processed and immediately discarded. We do not use submitted content for training, analytics, or any other purpose.

Methodology

How We Test Our Humanizer

Every update goes through the same four-step process before it reaches users. No exceptions.

Generate a baseline essay

We produce a 500-word essay using GPT-4o with no additional instructions — a clean, unmodified AI output that represents the default detector target.

Record detector scores before

We run the raw essay through GPTZero, Turnitin AI detection, and Copyleaks and log each score individually. This is the baseline we need to beat.

Humanize using Standard mode

The essay goes through UnAI My Text — no manual editing, no cherry-picking outputs. We use the same tool version that’s live for users at the time of the test.

Re-run all three detectors

We record pass or fail on each platform and track the scores over time. Tests run on the same day each month so conditions stay consistent across updates.

96–98% pass rate across GPTZero, Turnitin, and Copyleaks Current result — updated monthly. Last tested June 2026.

“

No lab. No patent. Just a working tool.

We’re a small team. We don’t have a research lab or a patent. What we have is a working tool, a clear testing methodology, and a commitment to keeping it free. We won’t put fake founder photos on this page or list credentials we don’t hold. We’d rather you judge us by what the tool actually does.

Common questions

FAQ

Straightforward answers to things people actually ask us.

Yes. The core humanizer is free and we intend to keep it that way. We may add optional features in the future, but the ability to humanize text will always be available without a subscription or credit card.

No. Your text is processed in real time and discarded immediately after the response is returned. We do not log submissions, store outputs, or use your content in any way after your session ends.

A small independent team. We’re not a VC-backed startup and we don’t have a PR department. We built this tool because we needed it and couldn’t find one that worked well without a paywall. That’s the honest answer.

We run our testing cycle monthly and ship improvements whenever the data shows a meaningful gain. If a major detector update changes the landscape significantly, we prioritize an out-of-cycle fix rather than waiting for the regular schedule.

The tool is tested against Turnitin AI detection and consistently passes. However, institutional policies on AI use vary widely — it’s your responsibility to understand what your school or program allows before submitting any work.