The MMI tool built by the tutor behind 72 offers and a 93%+ interview success rate.
150 ethics, role-play, conflict and policy stations. Speak your response or type it. AI feedback scored across Empathy, Communication, Reasoning, Reflection, Real-world Awareness, and Problem Solving. First review free.
Practice now jumps straight into any of the 200 stations with free AI feedback. Or take the Baseline Mock: 8 stations hand-written by Dan across every interview theme, for a true read on where you stand and a score you can track over time. How the Baseline Mock works
This is the feedback you get on every answer.
A complete Premium review, end to end - the marking on the criteria examiners actually use, plus the delivery, presence and rewrite layers. Nothing here is abridged, and your first review is a free Premium review exactly like this one.
/ 5
Your answer does not lean on rehearsed structure to hide thin reasoning - the gaps above are timing and empathy, not authenticity. When an answer sounds fluent but says little, this flag trips and the review shows you exactly where.
Your instinct to be useful is real, and examiners want future doctors who act. The fix is not to care less - it is to time it.
You moved to fixing before the person felt heard. Candidates in the top band sit in the emotion for a beat first, then problem-solve.
"Failing one subject isn't the end of the world, and lots of people fail and still graduate."
"That sounds crushing - especially tonight, when you were finally going to breathe. Before we work anything out, I just want to be here with you for a minute."
It names the feeling, signals you are staying, and earns you the right to problem-solve next - so your resit/appeal advice lands instead of bouncing off.
How you came across
Every Premium review measures this from your actual recording. Examiners score substance, not polish - this panel exists to help you sound like your most composed self.
- Swap fillers for a brief silent pause - it reads as considered, not hesitant.
- Your longest pause was 19.3s - fine for thinking, but narrate it ("let me think about that for a second") so it does not read as freezing.
Natural beats fluent. Examiners never mark down fillers or pauses, and chasing these numbers until you no longer sound like yourself will cost you far more than a few stray fillers ever could. This panel helps you notice habits, never a target to perform to.
Held the lens well through your opening. It drifted down-left each time you listed options - if you use notes, glance and return rather than reading.
Upright with a slight lean-in while talking about your friend - reads as genuinely involved rather than performing concern.
Warm at the start, then your face went neutral once you moved into logistics. Let the concern stay visible while you problem-solve.
You touched your face repeatedly in the second half - invisible to you, distracting on camera. Anchor your hands and it disappears.
Read gently, never to penalise a natural, cultural or neurodivergent style - this exists so nothing about your manner surprises you on interview day.
"That sounds crushing - you have worked all year, and to find out tonight of all nights. I'm not going anywhere, tell me what happened. ... Feeling like giving up makes sense after news like this, and I want to check - is it just the subject, or is everything feeling too heavy right now? ... When you're ready, we can look at options together: most unis allow a resit or an appeal, and we could draft the email to your coordinator tomorrow. I do have my exam in the morning, so here is what I can do: I'll stay on the phone as long as you need tonight, and I'll call you straight after my exam. You will not deal with this alone."
Re-record this station. For the first 20 seconds, say nothing that contains a solution - only what you notice they are feeling. Then compare your delivery read: steadier pace, fewer fillers, and the warmth audible.
This is the feedback on every answer.
Record a response and get this back - your first review is a free full Premium review, delivery and presence read included.
Practise a station free →Answer one MMI question. Get marked in 60 seconds.
The same AI examiner behind the practice tool. Speak or type your answer to the station below and see how it reads you - right here, before you sign up for anything.
Your close friend has started selling detailed notes for a course you both take, including material the lecturer explicitly asked students not to share. They tell you they are saving to help their family through an emergency, and they ask you not to say anything.
What would you do, and what would you weigh up?
One taster per person. A quick read, not a full review - your first full Premium review is free with an account. Answers are marked and not stored.
Practice MMI in three steps
No download or special equipment required. You can browse all stations before creating an account.
A coach that learns the mistakes you keep making
Turn your coach on and it reads your own AI-marked answers, works out the mistakes that keep coming back, and steers every session there. This is not generic feedback. It gets more tailored to you the more you practise.
What is in the MMI station bank
Stations are written to match the format and difficulty of real Australian medical school MMI circuits, including ethics dilemmas that require structured moral reasoning and role-play scenarios that test communication under pressure.
Plus the stations you write yourself. Heard a question in an interview group chat, or worried about one specific scenario? Write it as a station, or describe it roughly and have it built into a full station with five escalating questions, then practise it with the same timing, recording and marking as the bank. Stations you write stay private to your account. Included with any Key2MD subscription.
Five MMI timing modes
Real Australian MMI circuits give you roughly 2 minutes to read a station and about 8 minutes at the door. The practice tool lets you rehearse that exact rhythm, or drill faster and harder to build instinct. Prompts are revealed progressively, just like a live circuit.
| Mode | Reading | Speaking | Best for |
|---|---|---|---|
| Standard | 2 min | 5 min (3 prompts) | Realistic single-station conditions |
| Extended | 2 min | 8 min, advance prompts freely | Building full, structured answers |
| Quickfire | 45 sec | 15 sec read + 60 sec each (4 prompts) | Rapid recall and fluency under pressure |
| Random Probe | 2 min | Hidden timer, random follow-ups | Interview-style pressure and unpredictability |
| Custom | Your choice | Your choice | Targeted drilling of specific weak spots |
Six MMI criteria scored 1-5
Each criterion is scored from 1 (not demonstrated) to 5 (strongly demonstrated) with written feedback on why the score was given and how to improve it.
| Criterion | What the AI assesses |
|---|---|
| Empathy | Does the response demonstrate understanding of the perspectives and feelings of everyone involved? |
| Communication | Is the response structured, clear, and appropriate in tone for the context? |
| Reasoning | Is the ethical or practical reasoning sound, developed, and applied to the specific scenario? |
| Reflection | Does the response show self-awareness, acknowledge uncertainty, and avoid overconfidence? |
| Real-world Awareness | Does the response show understanding of how healthcare systems, resources, and constraints work in practice? |
| Problem Solving | Faced with an unfamiliar problem, does the response frame what is actually going on, compare options against real constraints, and commit to a workable plan? |
Alongside the six criteria, every response gets a single overall examiner mark out of 5 with a band - Poor, Unsatisfactory, Satisfactory, Good or Excellent - formed the way a real panel forms it: a holistic judgement that weighs genuineness and empathy most heavily, balanced against the strength of your reasoning. It is not just an average of the six criteria.
Because the panel marks a holistic impression, delivery carries real weight. Your tone, your prosody (the rise, fall and stress in your voice), your rhythm and pace, and the presence and composure you project decide whether the same words land as warm and convincing or flat and rehearsed. Premium reviews read these voice and presentation signals directly; spoken and typed modes sharpen the structure and substance they sit on.
For a full explanation of how each criterion is scored and what a strong response looks like, see how the AI marks your responses.
This is a coaching benchmark to track your improvement across sessions, not an official interview score. No medical school ever sees it.
Practise without a camera
The typed mode lets you write your response instead of speaking it. You receive the same six-criteria AI feedback as the spoken mode. This is useful when you want to practise response structure and argument quality without needing a microphone or quiet space.
More than a question bank
Reading prompts off a list is not what interview day feels like. The tool gives you the pressure, the follow-ups and the progress tracking a real MMI prep program needs.
Recordings are also saved on your device as you speak - if your browser ever crashes mid-station, your answer is offered back to you when you return, not lost.
When you are ready, sit the real thing
Two paid mocks bookend your prep. The Baseline Mock tells you exactly where you stand when you begin. The Full MMI Mock is the dress rehearsal for the week before your interview: eight stations you have never seen, hand-written by Dan and kept off the practice bank, run back to back with real timing and no feedback until the end.
MMI Pro subscribers get 20% off both mocks automatically. Launch code MMIMOCK15 takes 15% off the Full Mock until 19 July 2026, and code BASELINE15 takes 15% off the Baseline Mock until 31 July 2026 (the better single discount applies, they do not stack). Mock reports indicate practice performance only - they do not predict or guarantee your real interview outcome.
Practise live with Dan, every Friday at 6pm.
The AI tool builds your reps between sessions. The weekly live MMI classes are where you perform under real observation: small-group classes on Zoom where volunteers answer real interview questions and hear them marked and discussed on the spot. Performing is always optional, the questions are relevant to every medical school, and the skills carry into every interview format.
First review free, then $14 per credit
The station bank is free to access without an account. AI review feedback requires a signed-in account.
MMI practice FAQ
200 stations. AI feedback on six criteria. Speak or type your response. First review free.