We trained a humanizer on 2,590 real long-form essays.

What is the 2,590-essay corpus?

The research corpus contains 2,590 long-form English essays, over 5 million words of real writing, collected with permission. The humanizer is trained on it, so when you paste an AI draft it does not guess what natural writing looks like; it rewrites toward the cadence of writers who were actually doing the work, while your meaning, claims and facts stay fixed.

We publish the corpus methodology because nobody else in this niche does. QuillBot trains on "tens of thousands of human texts" without disclosure. WriteHuman ships a closed model. We tell you exactly what shaped our humanizer because that transparency is the moat.

How does HumanizeMyAI evaluate its output?

HumanizeMyAI evaluates every model change against dated internal evaluation sets that span several writing domains, scoring meaning preservation, readability, and detector variance. On 31 August 2026 our output read 0% AI on GPTZero, Copyleaks and the QuillBot detector, 0 to 3% on ZeroGPT, and Human on Turnitin and Originality AI. The dated screenshots are in the September 2026 product update, with earlier runs in the May 2026 model update and the vs WriteHuman head-to-head.

Every AI detector request goes through the same trained model we run ourselves, not a thinner public version of it. The score on screen is the score the engine produced.

Who built HumanizeMyAI?

HumanizeMyAI is built and operated by Fırat Mıhcı, an ESL writing researcher. Earlier published work on second- language writing patterns shaped the corpus selection criteria: every essay in the training set was chosen because it exhibits the specific structural patterns detectors look for in human prose. The product is a research project that became a product, not a product retrofitted with a research narrative.

There is no marketing team, no editorial board of fake personas, no "AI-powered customer success" chatbot. The address on the contact page reaches Fırat directly.

Why is there a free tier?

The free tier exists so the methodology claim is testable before any money moves. Opening an account gets you four humanizations of 250 words each, once, with no card and no growth-hack funnel behind it. Paste your own AI draft, run it through, then run our output through any review tools used in your workflow. Pay only if the result is what you needed. We're confident enough in the output to make the test free.

What does HumanizeMyAI refuse to do?

Four things this category treats as normal are off the table here: commission on the tools we review, training rights over the text you paste, invented author personas, and features whose point is misrepresenting authorship. Each one is a decision we can be held to, so each one is written down.

• We don't take affiliate revenue from any competitor we review. The QuillBot and Duey reviews are commissionless.
• We don't train our model on your inputs. Your run history keeps the text you sent and the text we returned for 12 months so you can find it again; none of it becomes training data.
• We don't use fake editorial personas. Fırat is the byline; future contributors will be real, LinkedIn-verifiable people, not stock-photo writers.
• We don't ship features that compete with academic integrity policies. The product is for AI-assisted writing improvement, not for misrepresenting authorship.

Is HumanizeMyAI any good?

HumanizeMyAI is built for one job: rewriting AI-assisted drafts so they carry the cadence of the human writing they were trained on. The evidence behind that is the corpus, 2,590 real essays and more than 5 million words, published rather than described. Every claim below is one you can settle yourself before you pay us anything.

Start with the corpus, which is published, so the training claim is inspectable rather than asserted. Then use the free tier: it asks for an account but never a card, so your own writing is the test case. Our detector figures are measurements with the date they were taken printed beside them, which is what makes them repeatable on your own text rather than a screenshot you have to trust.

The same lab publishes open research on how detection actually works, which is why the claims here carry dates and methods instead of adjectives. Run a paragraph of your own through the tool and then through a detector nobody here controls. That check settles the question faster than any page can.

What does the research lab publish?

The research lab publishes open computational-linguistics work on AI writing style, detector behavior, and second-language writing, each study shipped with its method, its code, and a DOI attached, in the Computational Linguistics Lab. One of those studies, a 12,222-review analysis for the Alanya City Council, ran in local newspapers; independent interviews with the founder and coverage of the lab's work are collected (with links out to each original piece) on the Press & Media page. Everything is published under one verifiable identity: Google Scholar, ORCID and ResearchGate.

What is next on the roadmap?

The roadmap is short and checkable rather than visionary. Next: open the eval set as a downloadable CSV, ship the detector matrix as a public benchmark page, and onboard one or two real freelance researchers as bylined contributors on non-technical pieces. The goal is transparency at every layer: methodology, data, code, authorship.

Try the humanizer · Pricing · Research · Press · Contact