Data-Driven, Not Wishful Thinking
Enter your GPA, test scores, and extracurriculars to see true empirical probabilities calibrated on 1,122 real applicant results and 6,273 US colleges.
AdmitGPT calculates your actual acceptance probabilities from your GPA, test scores, and extracurricular profile across thousands of colleges. Ground-truth mathematics, zero paywalls, and complete on-device privacy.

Ready to test your profile? Enter your academics and activities, and AdmitGPT will calculate your reach, target, and safety ranges instantly.
Rates reflect verified IPEDS institutional statistics. Your personalized odds depend on your academic z-score, extracurricular spike, and major competition.
Enter your GPA, test scores, and extracurriculars to see true empirical probabilities calibrated on 1,122 real applicant results and 6,273 US colleges.
We explain exactly how your spike score, academic z-score, and major selectivity impact your chances. No black boxes or paywalled secrets.
Every calculation runs directly in your browser JavaScript engine. No account required, no cookies sold, and zero data leaves your machine.
"I am like you — a student. I built AdmitGPT because I faced the same confusing black box you are facing now. The mathematics behind admissions shouldn't be a trade secret sold for thousands of dollars. We made it transparent, rigorously calibrated, and free for all students."
Ziad Salah — Creator of AdmitGPTTakes about 90 seconds. Enter your profile and see where you actually stand across your entire college list.
AdmitGPT is a free, open-source AI-powered college admissions probability calculator that gives you honest, data-driven estimates for over 6,000 US colleges. Unlike paid consultants or black-box models, every formula is published and auditable. Enter your GPA, SAT/ACT scores, and extracurricular profile to see your personalized admission chances for Ivy League schools, top national universities, liberal arts colleges, and state flagships.
A college admissions chance calculator estimates your probability of acceptance at a given school from your academic and extracurricular profile. AdmitGPT does this with a transparent additive-logistic model calibrated on 1,122 real applicant profiles and 6,273 colleges’ admission data (IPEDS). You enter your GPA, SAT/ACT scores, extracurriculars, intended major, and early-decision plans, and it returns a percentage per school — ranked by how you compare to each college’s own admitted-student distribution. It is a directional planning signal, not a guarantee: at schools admitting under 10%, the model under-predicts the true rate, and its ordinal accuracy is AUC ~0.74 (it reliably ranks similar applicants in the right order).
The engine uses an additive-logistic model — the same class of model used in published admissions research — combining academic z-scores, a six-dimension extracurricular spike score, intended major fit, international context, and early decision timing. No hidden fees, no data collection, no account required. Everything runs in your browser.
Explore our library of data-driven guides to understand every factor in your application:
Most college chances calculators are black boxes — they give you a number with no explanation of how it was calculated. AdmitGPT publishes every coefficient, every weight, and every formula. The model is calibrated against real admissions data (AUC ~0.74), and its limitations are documented openly. It is built by a student, for students, and will always be free.
Whether you are aiming for Harvard, your state flagship, or somewhere in between, knowing your real odds helps you make smarter decisions about where to apply, when to apply, and where to invest your energy. No consultant can give you a more honest answer than the data itself.
No calculator can predict an individual outcome, and we are open about that. AdmitGPT is calibrated on 1,122 self-reported applicant profiles and validated to an ordinal AUC of ~0.74 — meaning it reliably ranks similar applicants in the correct order. Its percentage estimates are well-calibrated for schools admitting above 10–25%, but at the most selective schools it systematically under-predicts (predicted ~1.3% vs an observed ~24.6% in the hardest decile). We publish the full calibration analysis on the Transparency page and the underlying data on the dataset page.