The online assessment (OA) is the first technical gate at almost every quant firm — and statistically the deadliest: it eliminates more candidates than every later round combined, because it is automated, unforgiving, and format-specific. The formats cluster into five families.
The five OA families
- Mental-math sprints. High question count, brutal clock, often negative marking. The archetype is Optiver's 80-in-8 (80 questions, 8 minutes); SIG, IMC, and Flow Traders run relatives. Tests raw speed-accuracy; trainable in weeks with daily reps.
- Probability/logic MCQs. 10–30 questions, 20–60 minutes: expected value, conditional probability, sequences, estimation. The G-Research quiz is the famous hard case; most funds run something in this family.
- Coding OAs. HackerRank/CodeSignal DS&A sets (Citadel SWE-track, HRT, Akuna) — LeetCode-medium/hard under time, all tests must pass; see where LeetCode fits.
- Game-based assessments. Market-making sims, betting games, multi-round strategy games scoring your EV discipline and sizing, not just outcomes — Optiver, IMC, and Da Vinci lean on these.
- Proctored math tests. Camera-proctored, pen-and-paper style reasoning where you narrate your work (Akuna's math OA for QR is the known example) — testing derivation, not just answers.
Why good candidates fail OAs
Three repeating causes: format surprise (first contact with negative marking or the pace happens in the real test), no triage strategy (spending 90 seconds on question 3 of 80), and arithmetic rust (knowing the method but executing slowly). All three are fixable with format-specific practice — none is fixable by more textbook study, which is why strong students fail OAs at surprising rates.
Practice by format
Every firm's reported OA format is mapped in its funnel page, with 27 firm-specific OA guides and timed simulations: start from your target firm's page, take the timed sample cold to find your gap, then train the specific format daily for 2–3 weeks. For the mental-math family specifically, the 80-in-8 simulator plus a daily Zetamac warm-up (how they differ) is the standard regimen.
More career guides
Frequently asked questions
What is a quant online assessment?
The automated first technical gate at quant firms: timed tests in one of five formats — mental-math sprints (Optiver's 80-in-8), probability/logic MCQs (G-Research's quiz), coding assessments (HackerRank/CodeSignal), game-based assessments, and proctored math tests.
Why do strong candidates fail quant OAs?
Format surprise (first contact with the pace or negative marking during the real test), no triage strategy, and slow arithmetic execution. All are fixed by format-specific timed practice rather than more textbook study.
How long should I prepare for an OA?
For the mental-math family, 2–3 weeks of daily timed practice moves scores materially. Take a timed simulation cold first to measure the gap, then train the specific format — pace, marking scheme, and question mix — of your target firm.
Do all quant firms use the same OA?
No — formats differ sharply by firm and role track (a Citadel SWE OA is coding; an Optiver trading OA is mental math and games; an Akuna QR OA includes proctored math). Check the specific firm's reported format before training.
Practice the real thing
QuantVault has 2,800+ quant interview problems with full solutions, intuition, and hints, firm-by-firm interview funnels, and an auto-graded coding judge. Start free.