CASPer prep courses compared honestly means comparing approaches, not brands. There are five: the free official practice from the test maker, large prep companies staffed by contractor coaches, AI practice platforms, free peer practice groups, and a single coach who has sat the test. Each is good at something real. Start with the official free practice test regardless of what else you do, because Acuity Insights states that applicants who complete it tend to perform better, and it costs nothing. Then add whichever of the other four fixes your actual gap.
What is each approach genuinely good at?
Official practice from the test maker. Acuity Insights gives every registered applicant a free full 11-scenario practice test inside their account, which it states simulates the exact experience and format of the real test, and which you can repeat as often as you like. Nothing else reproduces the interface, the countdown or the camera as faithfully, because nothing else is the interface. It is also the cheapest thing on this page. Acuity states plainly that it is not affiliated with third-party courses, does not endorse them, and that its data shows people using third-party materials do worse than people using its free resources. You should weigh that against everything below, including my own offer.
What it does not do: the practice test does not save or mark your answers. It tells you how the test feels, not how you did.
Large prep companies with contractor coaches. Their strength is scale and structure. A big bank of scenarios, a defined curriculum, scheduling across time zones, and availability when you want a session on a Tuesday at 9pm. If you need a syllabus and a lot of material, this is the category that has it. The structural trade-off is that the person who sells the programme is usually not the person who coaches you, and the coach may change between sessions.
AI practice platforms. Strong on volume and turnaround. You can do twenty scenarios in a week and get instant commentary on each, at any hour, which is genuinely hard to replicate with a human. Good for fluency, for getting your typing up to speed, and for breaking the blank-page problem. Their weakness is calibration: a model will assess structure reliably and depth unreliably, and it tends to approve of answers that are well-organised and empty.
Free peer practice. Two or three applicants, a timer, and a shared document. Excellent for the mechanical parts: typing under the clock, not overrunning, getting used to writing something rather than nothing. It is free, it builds a habit, and the accountability of a standing appointment is underrated. Its limit is that nobody in the group has been rated, so a group can converge confidently on a mediocre house style.
A personal coach who has sat the test. The strength is that one person sees your whole arc, remembers what you did wrong in week one, and can tell you whether an answer is hollow rather than whether it is tidy. Someone who has sat the test knows what the time pressure does to your judgment, which is different from knowing what the rubric says. The limit is capacity: one person has a finite number of hours, and if you want a session tomorrow, you may not get it.
How do you choose between them?
By naming your actual gap rather than your anxiety.
- You do not know what the test is like: free official practice test, twice.
- You freeze at the blank page or type too slowly: peer group or an AI platform, high volume, low cost.
- You have no plan and no material: a structured course.
- Your answers are organised and still landing in the middle: a human who can tell you why. That is a feedback problem and volume will not fix it.
Most applicants buy for the third gap and actually have the fourth.
What do I do differently?
I sat CASPer four times and scored in the 4th quartile each time, and I teach what I did rather than a rubric I read about. Three specifics:
- The same person coaches you throughout. There are no contractors. If you book a session, it is with me, and the person marking your week-four answer is the person who marked your week-one answer.
- You see scored examples. Not just principles: answers, with the marking reasoning next to them, including weak ones.
- The 4th Quartile Promise. If you do not reach the 4th quartile, we keep working together next cycle at no extra cost, until you do. It applies to students who finish the programme, every module watched, every assigned scenario submitted, every booked session attended. It covers my coaching, not your test fee, and it applies to one programme at a time. CASPer only.
The offer, with prices:
- Starter Success Package. The complete CASPer course, the practice scenario bank, the test-day sheet, the trained AI model for feedback at any hour, and check-ins from me as you work through it.
- Pro Prep Package. Everything in Starter plus two 60-minute private sessions with me: live mock tests, marked with personal feedback, recorded for you to keep.
- Elite 4th Quartile Package. Everything in Starter plus a private 60-minute session with me every week until your test day, with a fresh marked mock each week.
All three are on the packages page. If you are not sure which gap you have, a free mock exam marked by me is the fastest way to find out, and I would rather tell you to use the free practice test than sell you something you do not need.
A worked example of what feedback has to catch
An original scenario, written for this page:
You are a teaching assistant. A student emails you at 11pm asking for a two-day extension because his grandmother died. You know from a classmate that he told a different instructor the same thing last term. What do you do, and why?
Weak answer, and note that it is well structured. “First, I would acknowledge the student’s situation and express sympathy for his loss. Second, I would consider the fairness of granting an extension relative to other students. Third, I would consult the course policy and my supervisor. Finally, I would respond to the student. This approach balances compassion with fairness and ensures I follow appropriate procedures.”
A structural checker would pass that. It has an order, it names compassion and fairness, it escalates. It is also empty: it contains no decision, no reason, and no handling of the one fact that makes the question hard.
Strong answer. “The thing that makes this difficult is secondhand information, so I would deal with that first. What I have is a classmate’s account of something a student said to a different instructor last term. That is hearsay, I have not verified it, and acting on it would mean punishing a grieving student on the strength of gossip. So it does not go into my reply tonight.
Tonight, I grant the extension or I apply whatever the course policy says, and I reply briefly and kindly. If the policy requires documentation, I say so in the same way I would to any student, without the implication that I am asking him specifically because I doubt him.
Separately, and not tonight, there are two real questions. One is whether our course policy requires me to verify anything at all, because if it does not, I should not be deciding case by case on instinct. The other is what, if anything, I should do with what the classmate told me. My view is that I should not carry a rumour to the instructor or to the department, but I would tell the course coordinator that I am uncomfortable making these calls alone and ask that requests like this go through a single documented process. That fixes the pattern, if there is one, without me accusing anyone of anything I cannot support.
What I would not do is quietly refuse the extension while telling myself it is about fairness. If I have decided he is lying, I owe him that to his face with evidence, and I do not have evidence.“
Why the second one scores and the first does not. The second one identifies the actual tension, which is acting on unverified information, and resolves it explicitly. It separates what happens tonight from what happens afterwards. It names the option it rejects and the self-deception involved in it. It protects the student’s dignity while still addressing the systemic problem. None of that comes from a better template. It comes from someone telling you that your tidy four-step answer was empty, which is the one thing a checklist, a bot and a peer group all struggle to do.
Key numbers
- 11 scenarios in the free official practice test, repeatable as often as you like, at no cost.
- 0: the amount of feedback the official practice test gives, because answers are not saved or marked.
- 3: the CASPer packages on this site, all on the packages page.
- 1 free mock exam, marked by me, if you want help naming the gap before you spend anything.
- 4: the number of times I have sat CASPer and scored in the 4th quartile.
If you want the course and the feedback in one place, start here: the CASPer programme.
Related: Is CASPer prep worth it? and How to find a CASPer tutor.
Last reviewed 17 September 2026. Written by Arvin Ardakani, final-year medical student, 4th quartile on CASPer four times.