Online assessments

Quant online assessment practice: every firm's OA in one place

One hub for every online-assessment surface on QuantVault, grouped by firm: 52 firms, 40 timed OA replicas (each with a Pro full run), 17 free game replicas and practice tests, and 51 per-firm OA guides. Each row says what is free and what is Pro, and launches straight into the timed player, the game replica or the guide. Your last score on each timed replica shows up here once you have played it.

Play everything All Optiver OA games free SIG Quantitative Evaluation practice free Optiver 80 in 8 free 2027 OA census: all 42 firms Platform by firm: SHL, HackerRank, Codility, Criteria, Mettl
Showing 148 of 148 surfaces across 52 firms

Timed OA replicas (free sample + Pro full run)

Each timed replica reproduces one firm's reported online assessment sitting: the same question count, the same clock and the same answer format the real platform uses, whether that is numeric entry, multiple choice, forced-choice triads or a Python editor for coding screens. The free sample runs the opening question plus one from the harder back half so you can feel the pace with no signup; the Pro full run plays the complete sitting and grades it against our answer key. Your last score on each replica is saved in this browser and shown on its row.

Akuna Capital

4 free + 3 Pro to play
Timed replica
OA guide Interview process

JPMorgan Chase

3 free + 3 Pro to play
Timed replica
OA guide Interview process

Optiver

10 free + 2 Pro to play
Timed replica
OA guide Interview process

Susquehanna (SIG)

5 free + 4 Pro to play
Timed replica
OA guide Interview process

Free game replicas and drills

These are the game-style and mental-math screens that firms run in the browser before any interview, rebuilt as free runs with no account: Optiver's 80 in 8 and its Zap-N cognitive games, CCAT-style aptitude batteries, fill-in math screens and number-sequence tests. Most of them generate fresh questions from parametric templates on every run, so you can retake them as often as you like. Scores and leaderboards, where a game has them, stay on the game page.

OA guides (no replica yet)

Firms in this section have a written OA guide but no playable replica yet. Each guide covers the format, platform, timing and tracks that candidates report, plus the problem types worth drilling, and the card shows the platform and timing from our 2027 OA census. Replicas are added as new sittings are reported, so a firm can move up into the timed section between cycles.

Goldman Sachs

HackerRank / 52 min: Software Fundamentals 30 min (1 DSA problem) + Math 22 min (8 MCQ), proctored (reported)
Guide
What it is: Engineering New Analyst (Asia Pacific, 2027): one data structures and algorithms problem, then 8 multiple-choice questions on calculus, linear algebra and probability
Interview process

Man Group / AHL

5 parts incl. 50 number patterns / 10 min and about 10 to 12 probability / 12 min (reported)
Guide
What it is: Quant: table arithmetic, number patterns, bank-style logical reasoning, an array coding task and a probability block; platform possibly Correlation One (unconfirmed)
Interview process

Maven Securities

50 mental math / 5 min + 30 sequences / 6 min + 18 prob / 30 min (reported)
Guide
What it is: Trading: numerical (mental maths + sequences) + probability, wrong answers penalised; some tracks: Arctic Shores games; tech: HackerRank; 2026 variant adds typed answers + 80% CI estimation
Interview process

Shell

untimed work-style battery (4 parts, 84 items, resumable) + 3-question one-way video (3 min each)
Guide
What it is: Early careers: judgment scenarios (most and least likely), an information-security work simulation, self-ratings and 55 statement pairs; no maths screen
Interview process

Questions people ask about quant OAs

Is the CTC online assessment automated?

Yes. Our CTC OA guide describes a timed online screen taken before any interview: Quant Trading Analyst candidates get an aptitude assessment of logical word problems, mental math, estimation and probability questions under tight time pressure (one account cites 34 minutes), while Software Engineering candidates get a Codility coding challenge with a small number of data-structures-and-algorithms problems. Rehearse the aptitude format on the free CTC aptitude test replica and the SHL-style reasoning battery in the timed CTC replica above.

How long is the SIG Quantitative Evaluation?

17 questions in 60 minutes on the Mettl platform, calculator allowed, for Trader and Quant Researcher candidates. It covers probability, expected value, Bayes, constraint logic and gambler's ruin, and the bank has recurred since 2019 with only the numbers changing between sittings. Try three free questions on the SIG practice page; our full simulation runs the whole recurring bank, 22 questions in 60 minutes.

Which quant firms use SHL assessments?

Four firms on this page report SHL or SHL-style screens. Barclays runs an SHL interactive cognitive test of 8 activity-based questions in 12 minutes (drag, drop, tap and arrange rather than multiple choice). IMC pairs an SHL-style cognitive assessment with an untimed SHL Values and Mindset questionnaire of 76 forced-choice triads. JPMorgan's 2027 campus candidates report the same 76-triad SHL Values and Mindset questionnaire alongside the coding screen. CTC's aptitude test is described as an SHL-style reasoning battery. The Barclays and JPMorgan screens have timed replicas above, and IMC's values test has its own free replica.

Which quant OAs run on HackerRank?

HackerRank is the most common host in our 2027 census, 14 of 51 firms: Akuna Capital: 3 problems (coding); 30 fills in ~12 min (trader); Arrowstreet Capital: 11 numeric + MCQ block / ~4 min per question; Citadel / Citadel Securities: 2 problems / 70 min; Hudson River Trading: 3 problems / ~150 min; IMC Trading: 15 questions (QT) / 11 questions (QR); Jane Street: ~4 questions / 30 min; JPMorgan (NAMR Quantitative Research): 2 problems / 60 min; Point72 / Cubist: 5 questions; Qube Research & Technologies: 3 problems; Squarepoint Capital: ~4 MCQ + 2 coding problems; Two Sigma: 3 problems / ~3 h; WorldQuant; Morgan Stanley: 120 min: 12 MCQ (6 statistics + 6 applied math) + 2 Python tasks (IED Quant Finance, HK); Goldman Sachs: 52 min: Software Fundamentals 30 min (1 DSA problem) + Math 22 min (8 MCQ), proctored (reported). Two examples with replicas above: JPMorgan's NAMR Quantitative Research screen is two Python questions in 60 minutes with a timer that cannot be paused, and Akuna's 2027 quant coding challenge is three HackerRank Python questions in 120 minutes with fullscreen forced and a multiple-monitor check. Akuna Quant Researchers also sit a separate camera-proctored math OA on EasyHire, where you verbalize your reasoning on paper.

Which platform does Millennium use?

It depends on the track. Some candidates report a Criteria assessment early in the funnel: the Criteria Cognitive Aptitude Test (CCAT), 50 questions in 15 minutes, no calculator and no penalty for wrong answers, which you can rehearse on the free Millennium Criteria replica. Quant Researcher interns get a 24-hour take-home market-making project in a Jupyter notebook (four challenges), plus a Caliper personality test, and an earlier QR cycle used a 120-minute, 7-question HackerRank test. Quant Developer, Data Engineer and Data Scientist tracks each get their own HackerRank battery; see the Millennium OA guide.

Can you retake a quant online assessment?

Our firm guides do not document a retake policy for any firm, so plan for one real sitting and treat the practice here as the retakes. Every game replica on this page runs again with fresh content (the CCAT-style tests build new items in your browser on each run and the Akuna boards reshuffle every run), and each timed replica stores your last score in this browser, so you can see whether a second attempt improved it. What our census does track is whether the real bank repeats between sittings, covered in the next answer.

Do the OA question banks repeat between sittings?

Often. Our 2027 census marks 18 of 51 firms as recurring banks (Akuna Capital, AQR Capital Management, Belvedere Trading, Bridgewater Associates, Citadel / Citadel Securities, D. E. Shaw, DRW, Hudson River Trading, IMC Trading, JPMorgan (NAMR Quantitative Research), Jump Trading, Old Mission Capital, Optiver, Point72 / Cubist, SIG (Susquehanna International Group), Squarepoint Capital, Two Sigma and Virtu Financial) and 2 as rotating (G-Research and XTX Markets); the rest are blank because too few sittings have been reported. SIG's 17-question screen has recurred since 2019 with only the numbers changing, Akuna's 2027 coding pool showed only 7 distinct problems across four reported sittings, IMC's software screen draws two hard problems from a recurring bank, and JPMorgan draws its two questions from a small rotating pool. A recurring bank is why drilling the exact replica pays off.

Does a good Zetamac score matter?

As a pacing benchmark, yes; as a cutoff, no, because no firm publishes a Zetamac threshold. On default settings the candidate-reported bands in our Zetamac guide are: below 40 not ready to apply, 45 to 55 a workable prep baseline, 55 to 70 competitive for most prop trading firms, and 70 or more strong even for top market makers like Optiver or Jane Street. On the Optiver 80 in 8 itself, about 55 net is a commonly reported pass line and 70 or more is competitive at the top desks.

What does a free sample contain versus a Pro full run?

Every timed replica has a free sample that needs no signup: the opening question plus one from the harder back half, timed at the real per-question pace (the SHL values sample gives three of the 76 triads instead, and coding screens preview the first question read-only with its statement, sample case and editor template). The Pro full run plays the complete sitting on the real clock, for example Barclays: 8 questions in 12 minutes; IMC Trading: 15 questions in 30 minutes; JPMorgan Chase: 5 questions in 60 minutes; Susquehanna (SIG): 22 questions in 60 minutes; Akuna Capital: 8 questions in 120 minutes, grades it against our answer key and shows a full review; coding runs use a live in-browser Python editor with hidden tests. All 17 game replicas and all 51 firm guides are free.

Is the Optiver OA the same as the 80 in 8?

No. The 80 in 8 is a separate mental-arithmetic screen, 80 questions in 8 minutes with no calculator and +1/-1 scoring, and it sits alongside the main CodeSignal battery of 26 number sequences in 25 minutes plus 30 probability questions at 90 seconds each (the Quant Researcher track adds HackerRank). Together they form Optiver's front gate. Both are free here: the 80 in 8 replica and every Optiver OA game, including Beat the Odds, Number Logic, Likelihood List, Intervals, Order Books and the Zap-N cognitive games.

All firm OA guides (text index)