The Free Tools: Thirteen Addresses and What Grades

The thirteen free case interview tools, open without an account

The index at the tools page shows the calculators, the builders and the practice sets. The six graded drills each have their own address, so here is the complete set in one table.

Seven of those are calculators, builders and practice sets. The other six are the graded drills, and they are the reason the list is worth having in one place.

Six graded case interview drills, and the weights behind each

Every graded drill scores four dimensions out of 100. The weights below are ours, read from the grading code, and they are the numbers this site publishes because no firm publishes any. Bain goes furthest and says only that its feedback rests on "an objective scoring system".

Read the heaviest weight in each row and you have the whole teaching in six lines. Structuring lives on coverage, because a branch you never named cannot be reached later. Synthesis lives on position, because a recommendation is tested on your first sentence alone. Exhibits live on the so what, because a chart is worth what it changes. How an answer is assessed sets out the reasoning on each, and the same weights appear there because that page executes the same code.

Two of those weights answer a question candidates ask constantly. Units carry 15 on the math rubric and the sanity check carries 20 on market sizing. Both are habits rather than talents, which makes them the cheapest points anywhere in the set.

How the grading actually runs

Three mechanics are worth knowing, because they change how you use the drills.

1. It runs in your browser. The drill calls the grading function directly on the page, so no answer is sent anywhere and no account is involved. That is also why there is no history: each attempt is scored on its own text.
2. The clock starts on your first keystroke. Opening a drill and reading the prompt costs nothing, which matches a real case, where the count begins when you start talking.
3. An empty answer is refused rather than failed. A response with no number, no list and fewer than two sentences returns a single line saying there is nothing to assess, instead of a zero that reads like a wrong answer.

Scoring is deterministic. The same text always returns the same dimensions and the same notes, which is what makes a week over week comparison mean something. It also means the rubric checks whether the expected moves appear and cannot check whether your premise held. A well phrased answer resting on a bad assumption will score well here and fail in a room, and the method page says so in the platform's own words.

Three calculators that print the chain

The calculators are the tools most people underrate, because a calculator that only prints an answer is worth nothing in an interview. These print the steps, which is the part you have to say out loud. BCG names Show Your Thinking and Clarify Assumptions among six behaviors it wants in a case, and both are about the chain rather than the total.

Here is the strongest test of whether the break-even calculator does real work. Bain publishes a coffee shop case with every input: three pounds a cup, one pound of variable cost, 245,610 pounds of setup and 163,740 pounds of annual operating cost.

Add the two fixed costs and you get 409,350. Divide by the two pound contribution and you get 204,675 cups, which is the number the firm prints on its own page. The calculator returns the same, because the formula is the same. It also refuses to answer when price sits below variable cost, which is the case where a break-even would otherwise print as infinity.

The market size calculator runs the same idea on estimates: a chain top down or bottom up, then the total expressed as a share of the region's GDP and as spend per head. Nine geographies are preloaded, so a sanity check takes one glance rather than a second search.

Thirty drills, eight families, one seed

The math practice set is generated rather than hand written, from a seeded sequence, so the same thirty questions come back in the same order every time. It rotates through eight families: percentages, growth and CAGR, break-even, unit economics, weighted averages, return and payback, market sizing, and bare mental arithmetic.

Thirty across eight is an uneven rotation, and knowing how it falls tells you what you are actually drilling. Six families appear four times and two appear three times, because four sixes plus three twos is thirty. Answers are accepted within 1 percent of the expected value, with a floor of 0.05, which is tighter than the 2 percent the open response drill allows.

Eight worked estimates sit at the market sizing questions, covering a European port, office coffee in the United Kingdom, Spanish domestic flights, German heat pumps, a supermarket checkout lane, corporate language training in France, Italian bakeries and London lifts. Write your chain first, then open the worked one. Reading it first teaches almost nothing, since agreement feels the same as competence.

Where the case interview tools stop and an account starts

Said plainly, because a page that is vague about this is selling something. The thirteen tools above need nothing. The downloadable resources are different: the case books, the fit workbook, the resume templates and the trackers all sit behind an account, and the resources page says so on every card. Running a full case from the library also goes through checkout, and every plan carries the same core set.

The distinction that matters is between a paywall and a login. None of the thirteen tools above has either one. The resources have a login and the case library has a paywall, and confusing the two is how a page ends up promising a free download that turns out to want a card. If a link on this site asks you to sign up before showing you something the page already claimed, treat that as a defect on our side rather than a feature of a plan.

There is also no trial to cancel anywhere in the free set, because there is nothing to start. The drills have no session limit, no daily cap and no counter, since nothing is stored between attempts in the first place. That is a consequence of grading in the browser rather than a generosity, and it cuts both ways: you get unlimited attempts and you get no history, so keep your own one line log of which dimension you lost.

Bain writes that it is "not looking for a single right answer" and asks candidates to "Think aloud: Share your reasoning and calculations, even if you're unsure." Those two sentences describe what these tools train and also what they cannot: a drill has no interviewer to talk to. Use them for the dimensions and the habits. For the conversation, read our comparison with a private coach, which says where a human beats all of this outright.