Frequently asked questions

Short answers, with a link to the full guide where there is one. Still stuck? Email support@katabench.com.

Katabench is hands-on .NET practice. The puzzle catalog covers seven disciplines: algorithms, Database / EF, refactoring, architecture, secure coding, design principles, and test writing. Correctness is only one part of the evidence; applicable exercises can also grade time, allocations, query plans, code structure, design constraints, security behavior, or test effectiveness.

Katabench Labs extend that practice into guided multi-file projects. The How it works page is the product overview.

The System Design Studio adds a visual eighth practice surface. You build provider-neutral architectures on a canvas and receive deterministic feedback on paths, boundaries, redundancy, and failure isolation.

No. Puzzle editing and grading happen through the browser. Labs, when available, also provide their project workspace through the browser. You do not need a local SDK or Docker installation for the hosted experience.

Puzzle submissions use C# 12. Katabench is intentionally focused on the .NET ecosystem so its feedback can be specific. System Design challenges do not require a programming language; they use provider-neutral architecture components. See The editor & runtime for the language and library boundaries.

Yes. The Playground switches among Problem, Code, and Results views, and Labs switch between Guide and Code & output. A desktop is more comfortable for large multi-file katas, terminals, and detailed query plans.

Yes. Free requires no credit card and includes the full algorithms catalog, six System Design challenges, and the other features shown on Pricing. Plans, limits & billing explains how separate run, submission, and reveal allowances work.

Run executes visible samples and optional custom input. It does not use a submission. Submit grades the full suite, including hidden checks, and uses the submission allowance on Free. Run and Submit have separate daily allowances. Check is the System Design equivalent of Run: it evaluates the current topology without marking the challenge solved. See Getting started.

No. The pass verdict grades structural properties such as required paths, forbidden bypasses, redundancy, boundaries, and semantic connections. When the structure holds, a deterministic capacity simulation adds p99 latency, throughput, error rate, and modeled cost under the challenge's traffic scenarios and names the component that saturates first. Those numbers are a reproducible learning model, not a provisioned load test. See Using the System Design Studio.

Run proves only the visible examples. Submit adds hidden edge cases, larger inputs, and any track-specific quality gates. Open the Tests tab, use Only failing, and read the verdict at the top. Troubleshooting gives a diagnosis order.

Timings are measured in the grading environment around your code, not in the browser and not across your network connection. This makes puzzle speed standings comparable. Normal small timing variation still exists, so prioritize clear headroom.

No. Hidden tests show their name, status, timing, and applicable rule or budget evidence, but not the input or expected and actual values. Visible tests show the detailed difference you need for ordinary debugging.

No. Reveal is a study tool. It replaces the editor with a correct reference, uses the reveal allowance when applicable, and does not mark the puzzle solved. Read it, Reset, and reproduce the approach from memory.

Yes. Hints are unlocked one at a time, cost nothing, and do not reduce the reward for a solve. Only Reveal solution is metered, and a solve after a reveal earns the practice rate rather than the first-solve reward.

Only a clean graded Submit. Passing Run, opening every hint, or revealing the solution does not mark completion. One solve can also count toward every learning path that contains that puzzle.

A track is a discipline used to group similar puzzle types. A learning path is an ordered curriculum that selects exercises, groups them into sections, and tracks a finish line. You can solve the same exercise from either the path or the open catalog.

A puzzle is a focused graded coding task. A Lab is a guided project workspace containing ordered lessons, multiple files, application output or terminals, and an executable Check step. See Working through Katabench Labs.

Successful signed-in practice can award XP and gems and extend your streak. Daily challenges, quests, achievements, paths, and Labs can add rewards. The app displays the exact value and condition for each current reward. Gems buy titles, themes, and streak freezes. See Progress, XP, streaks & rewards.

Path progress uses solved puzzles. If you solved an exercise earlier from the catalog or another path, the matching step is already complete. This is intentional.

Pro includes the premium puzzle tracks, the remaining System Design curriculum, full hidden performance suites, every guided path, the Pro Lab catalog when Labs is available, and removes daily run, submission, and reveal caps under fair use. The feature matrix is the current source of truth.

Sign in, open Account & billing, and select Manage billing. The customer portal lets you update payment details, view invoices, or cancel. A cancellation keeps paid access through the date shown on the plan card.

See the Refund policy and Terms of Service. For an account-specific question, email support@katabench.com.

When signed in, solved and submission history, learning progress, Lab completion, XP, streaks, achievements, gems, badges, owned rewards, and your in-progress puzzle and Studio drafts belong to the account. Custom Run input and some view preferences are stored in the current browser.

Local editor buffers and browser preferences can be removed. Signed-in server-backed progress stays with your account. Guest-only progress can be lost, which is why creating an account matters before clearing data or changing devices.

Identity, email, and password are managed from the account menu in the header. To delete your account, use that menu or email privacy@katabench.com; the Privacy policy states how long data is retained afterwards and lists your other data rights.

Email support@katabench.com. Include the puzzle, path, course, Lab, or lesson name; what you selected; what you expected; what happened; and whether retrying changed the result. Do not include passwords or payment-card details.