Frequently asked questions

Short answers, with a link to the full story where there is one. Still stuck? Email [email protected].

Getting started

What is Katabench?

C# code puzzles graded the way production judges you, across seven disciplines: algorithms, SQL & EF Core, refactoring, architecture, secure coding, design principles, and test writing. Correctness is the floor; real time budgets, real allocations, real query plans, and real design rules are the bar. The How it works page is the full tour.

Do I need to install anything?

No. Everything runs in the browser — you write in the editor and the grading happens on our servers. There's nothing to download and nothing to configure.

Which languages can I use?

C# only, on purpose. The whole platform — the performance gates, the EF Core query-plan grading, the Roslyn-measured refactoring metrics — is built to judge C# the way a senior reviewer would. See The editor & runtime for the exact language version and libraries you get.

Is it really free?

Yes. The Free plan includes the entire algorithms track, performance-graded, with no credit card. Plans, quotas & billing explains what Free covers and what Pro adds; Pricing has the exact figures.

Grading & results

What's the difference between Run and Submit?

Run executes the visible samples and never counts against your submission quota, so you can iterate freely. Signed-in Free accounts do have a daily run ceiling of 50 as an abuse guard; Pro and Pro+ are uncapped. Submit grades you against the full hidden suite and counts toward your daily quota. The full distinction is in Getting started.

Why did Submit fail when Run passed?

Because the hidden suite tests cases the samples don't — edge cases and large inputs. It's the most common surprise on the platform, and Troubleshooting walks through exactly how to diagnose it.

Where do the timings come from?

From the grading servers, measured immediately around your code — never your browser or your network. That's what makes the speed leaderboards fair and comparable. More in How grading works.

Can I see the hidden test inputs?

No — hidden tests reveal only their name, pass/fail, and timing, so the suite can't be gamed. The test names are written to be honest hints. This is by design, not a limitation.

Is revealing the solution cheating?

No. Revealing loads a correct reference for you to study; it costs no submission and doesn't mark the puzzle solved. Read it, reset, and beat it from memory — that's deliberate practice.

Plans & billing

What does Pro unlock?

The four tracks beyond algorithms (Database / EF Core, Refactoring, Architecture, Secure Coding), the full hidden performance suites, every guided path, and no daily caps. The breakdown is in Plans, quotas & billing and the feature matrix.

What counts toward my daily quota?

Submissions, not runs. Running the samples never spends a submission; only grading against the hidden suite draws down the Free plan's daily allowance, which resets each day. (Runs do have a generous daily ceiling of 50 on signed-in Free accounts as an abuse guard, lifted on Pro and Pro+, but that's separate from your submission quota.)

How do I cancel or get a refund?

You can cancel any time and keep Pro until the period you've paid for ends. See the Refund policy for refunds and the Terms of Service for the full agreement, or email [email protected].

Account & data

Is my progress saved?

Your solved marks and in-progress code are kept per puzzle in your browser, so you can pick up where you left off. Clearing your browser's site data clears that local progress, so don't wipe it if you want to keep your buffers.

How do I report a bug or suggest a puzzle?

Email [email protected] with the puzzle name and what you saw. Good bug reports and puzzle ideas genuinely shape what we build next.