Geeky Medics is no longer merely an OSCE website, and anyone still filing it under free skills guides has missed one of the quieter transformations in UK medical education. Its AI Tutor can generate quizzes, flashcards and OSCE stations, analyse performance, identify weak areas and work from uploaded learning material; and alongside it sits a simulation layer of AI virtual patients, hundreds of scenarios, voice and text interaction, with AI examiner feedback (last checked August 2026). iatroX has taken a different road into the same future. This comparison maps both honestly, because they are genuinely complementary, and manufacturing a winner would misdescribe both.
What the Geeky Medics AI Tutor offers
The Tutor's shape is generative and learner-fed: conversational questioning, quiz and flashcard generation on demand, OSCE station generation, awareness of the learner's performance history, and the ability to ingest uploaded materials so practice reflects what your medical school actually taught. That last capability matters more than it sounds: undergraduate assessment is local, and a tutor that drills your lecture content is solving a problem national banks structurally cannot. The generative approach's trade-off is the classic one, flexibility and personalisation over item-by-item curation, which Geeky Medics manages sensibly by anchoring the AI in its large educational library.
The simulation advantage
Geeky Medics' genuinely distinctive asset is the virtual patient ecosystem: AI patients you take histories from, counsel and examine in structured scenarios, by voice or text, with an AI examiner scoring and giving feedback afterwards, repeatable, private, and available at 11pm in a way no simulation suite ever was. For communication skills, history-taking and OSCE performance under observation, this is the strongest AI offering in UK undergraduate education, and our roundup of that whole category is at /blog/best-ai-tools-for-osce-practice-virtual-patients-2026. Nothing in iatroX competes with it, by design.
iatroX's different emphasis
iatroX's centre of gravity is knowledge and reasoning across the exam-facing career: curated adaptive question banks for 40+ undergraduate and postgraduate examinations, the Socratic Tutor interrogating wrong answers until the underlying misconception is named and corrected against cited guidance, spaced repetition retesting the correction, the AI Study Planner sequencing it all, and askiatroX providing cited clinical reference when the context shifts from revision to practice. Where Geeky Medics generates practice from your materials, iatroX diagnoses and repairs your performance against the blueprint; where the virtual patient trains the consultation, the Socratic dialogue trains the thinking underneath it.
AI simulation versus AI tutoring
This is the educationally interesting distinction the two platforms embody. Simulation rehearses performance: the observable behaviours, structure, communication, examination sequence, that OSCEs assess, with feedback on what an examiner would see. Socratic tutoring rehearses cognition: hypothesis generation, discriminating features, what-would-change-your-mind, the invisible work that written exams sample and clinical practice runs on. Both are forms of deliberate practice; they target different layers of competence, and the learners who confuse them, drilling MCQs for an OSCE, rehearsing stations for an applied knowledge paper, are practising the wrong layer diligently.
Which approach for which learner
The undergraduate facing OSCEs: Geeky Medics is compelling, simulation plus the generative tutor over local materials. The finals or UKMLA candidate: both, honestly, stations through Geeky Medics, the AKT-style written paper through blueprint-mapped adaptive practice. The postgraduate facing MRCP, MRCGP or a specialty written exam: the requirements shift decisively toward curated, exam-specific banks with a diagnostic loop, iatroX's territory, with our exam-by-exam guides starting at /blog/best-ai-tutors-for-mrcgp-akt-and-sca. The clinician with day-to-day uncertainty: cited clinical reference, which is askiatroX's job and outside Geeky Medics' scope. Run as a pair across an undergraduate-to-postgraduate arc, the two platforms barely overlap, which is the most useful sentence in this comparison.
Frequently asked questions
Is the Geeky Medics AI Tutor free?
Geeky Medics maintains substantial free content, with the AI features sitting in its paid tiers; check current pricing on its site (last checked August 2026). Its pricing has historically been student-friendly relative to the category.
Can generated questions replace curated banks for finals?
For drilling and gap-exploration, generated items are genuinely useful; for calibration against a specific blueprint, curated and validated items remain the safer spine. The strongest preparation uses generation for breadth and curation for fidelity, which is exactly how these two platforms pair.
Does uploaded-material tutoring risk learning the notes rather than the medicine?
It is a real trade-off: drilling your school's slides optimises for your school's exam, which is rational locally and narrower than the national blueprint. The sensible pattern is uploads for the local assessment, blueprint-mapped banks for licensing exams, and awareness that the two can disagree, at which point the guideline, not either resource, is the arbiter.
Where does the Geeky Medics bank sit relative to its AI features?
Its question content and library remain the substrate the AI generates from, and for many undergraduates the free layer plus AI tier covers both breadth and rehearsal affordably. The comparison in this article is about emphasis, simulation-and-generation versus curation-and-interrogation, not about either platform lacking a foundation.