Cubiks Logiks vs Saville Swift: Two Fast UK Aptitude Tests, Same Shape
Cubiks Logiks and Saville Swift Aptitude are the two UK general aptitude tests that most resemble each other. Both cover numerical, verbal, and abstract reasoning in a single fast-paced session. Both are used heavily by UK graduate schemes and mid-to-large employers. The formats feel almost interchangeable until you notice that Saville gives you 6 extra minutes, splits 72 questions across 3 sub-tests, and weights abstract reasoning differently. Candidates often confuse invites between the two. Getting it wrong means prepping the wrong pace.
Start Free PracticeSide-by-side: Cubiks vs Saville
Near-identical on paper, different under the hood in section weighting and pacing philosophy.
| Cubiks | Saville | |
|---|---|---|
| Full Name | Cubiks Logiks General Ability | Saville Swift Aptitude |
| Vendor | Cubiks (PSI Services) | Saville Assessment |
| Questions | 50 (total) | 72 (total across 3 sub-tests) |
| Time Limit | 12 minutes | 18 minutes (6 per sub-test) |
| Seconds per Question | 14.4 seconds | 15 seconds |
| Sections | Numerical, Verbal, Abstract (interleaved) | Verbal, Numerical, Diagrammatic (separate sub-tests) |
| Adaptive | No | No |
| Calculator | Usually allowed (on-screen) | Usually allowed (on-screen) |
| Guessing Penalty | None | None |
| Scoring | Raw + percentile | Percentile per sub-test and composite |
| Headline Employers | BP, Shell, Santander, Aviva | HMRC, Mazars, Grant Thornton |
| Industry Lean | Energy, banking, insurance | Public sector, accounting, professional services |
| Feel | One fast sprint across 3 content types | Three short sprints, one content type each |
Format: one sprint or three short sprints
Cubiks Logiks General Ability runs 50 questions in 12 minutes as one continuous session, with numerical, verbal, and abstract questions interleaved. You do not know what the next question type will be until it appears, so you switch cognitive modes rapidly throughout. The interleaving is designed to test mental flexibility as much as raw reasoning ability. Cubiks also offers standalone Logiks Numerical, Verbal, and Advanced modules for employers who want deeper assessment in one area.
Saville Swift Aptitude runs 72 questions in 18 minutes, split into three sub-tests of 24 questions each (6 minutes per sub-test). Each sub-test covers one content type: verbal comprehension, numerical reasoning, and diagrammatic reasoning. The sub-tests are separate; you complete all 24 verbal questions before moving to numerical, and all 24 numerical before moving to diagrammatic. There is no interleaving within a sub-test.
The psychological and prep implications are meaningful. Cubiks rewards candidates who can switch cognitive modes fast (verbal item, then numerical, then abstract, with no warning). Saville rewards candidates who can sustain focus on one content type for 6 minutes before switching. These are different mental postures.
Timing: 14.4 seconds or 15 seconds
Cubiks gives you 14.4 seconds per question. This is tight, especially given the context-switching between interleaved question types. A numerical question followed by a verbal question followed by an abstract pattern question each require different cognitive modes, and the 14.4-second budget does not leave much room for the mental gear change.
Saville gives you 15 seconds per question, distributed as 15 seconds for each of 24 questions per sub-test. The per-sub-test structure means you settle into numerical reasoning for 6 minutes, get into the flow, then switch. Most candidates find Saville pacing slightly easier despite the minimally longer per-question budget because the sub-test structure eliminates context-switch cost.
Both tests share the experience of feeling rushed throughout. Fewer than 10 percent of candidates finish every question correctly. The realistic target on either is roughly 75 to 85 percent of questions attempted with 85 percent accuracy, which puts you in the 70th to 85th percentile depending on specific employer norms.
Question family comparison
The same three families. Different internal weightings.
Numerical reasoning
Cubiks: ~17 items covering percentages, ratios, chart reading, basic algebra. Saville Numerical sub-test: 24 items covering similar content with slightly more chart and table reading, less pure word problem. Both allow on-screen calculator.
Verbal reasoning
Cubiks: ~17 items with a mix of short-passage inference and vocabulary. Saville Verbal sub-test: 24 items almost exclusively passage-based with true/false/cannot-tell format. Saville verbal is dryer and more business-context heavy; Cubiks verbal is closer to classical SAT-style mixed verbal.
Abstract/Diagrammatic reasoning
Cubiks: ~16 abstract reasoning items (pattern series, shape relationships, matrix-style puzzles). Saville Diagrammatic sub-test: 24 items covering flow-chart-style diagrammatic reasoning, input/output rule discovery, and symbol substitution. Saville diagrammatic feels more business-operations flavored. Cubiks abstract is closer to classical pattern matrices.
Context switching (Cubiks-specific feature)
Cubiks interleaves question types. Saville does not. If you are easily thrown by having to switch from numerical to verbal unexpectedly, Saville is significantly easier for you. If you find you hit a rhythm faster by bouncing between content types, Cubiks is the friendlier format.
Which is actually harder
Subjectively, Cubiks Logiks feels harder because of the context-switching cost. Candidates report that the first 3 or 4 context switches feel jarring, and that accuracy drops on the first item of each new content type. Over 50 questions with ~15 switches, this compounds. Saville feels easier because you get to sink into one content type at a time.
Objectively, scoring percentile distributions are similar. Cubiks average raw scores sit around 30 of 50 (60th percentile threshold usually sits around 33 to 35). Saville average composite raw scores sit around 43 of 72 (roughly 60 percent accuracy), with percentile cutoffs typically at the 60th to 75th percentile for graduate roles.
The harder test for a given candidate depends on their cognitive profile. Candidates with uneven skill levels (strong numerical, weak verbal, for example) are exposed more on Saville because sub-test scores are often read individually by employers. Candidates with balanced skills do relatively better on Saville than on Cubiks. Cubiks' interleaving means a weak area bleeds into adjacent questions through context-switch drag.
Scoring and employer interpretation
Cubiks Logiks reports a raw score out of 50 plus percentile against the Cubiks graduate norm group. Most graduate employers set cutoffs at the 60th to 75th percentile (roughly 33 to 38 correct). Investment banking adjacent employers push higher (80th percentile or above). Sub-scores (numerical, verbal, abstract) are available but typically used only for diagnostic purposes unless the role specifically requires strong numerical, for example.
Saville Swift reports per sub-test percentile plus composite. Because sub-tests are separate, employers often set minimum percentile floors per sub-test. A typical graduate employer might require 60th percentile on each sub-test (roughly 14 of 24 per section) with composite at the 65th percentile. This per-sub-test floor catches uneven candidates more strictly than Cubiks single-score cutoffs.
Both tests have no guessing penalty, so the final seconds of either test should be used to fill every unattempted question with a best guess. For Saville specifically, this applies per sub-test because when a sub-test runs out of time, the remaining questions are locked. Always fill blanks at the end of each sub-test.
Where each test lives
Cubiks Logiks is used heavily in UK and European energy companies (BP, Shell, Equinor), banking (Santander, some UK retail banks), and insurance (Aviva, Zurich). Cubiks was acquired by PSI Services in 2020 which expanded enterprise distribution. Cubiks Logiks General Ability is the most common variant; standalone Logiks Numerical and Verbal appear in roles that specifically need deep assessment in one area.
Saville Assessment products are widely used by UK public sector (HMRC, Ministry of Defence graduate schemes), accounting firms (Mazars, Grant Thornton, BDO), and UK professional services generally. Saville has historically been favored by employers who value precise sub-test scoring and the ability to weight specific skills per role. It is less common in US or European tech hiring than Cubiks or SHL.
How prep differs in the details
For Cubiks Logiks, the highest-leverage prep skill is context-switching speed. Drill mixed-content practice sessions: 10 numerical items, 10 verbal items, 10 abstract items, all interleaved at the actual test pace (14.4 seconds). Track how accuracy drops on the first item after each content switch. Most candidates see 5 to 10 percentage-point accuracy drops on context-switch items. Specific interleaved drilling narrows that gap.
For Saville Swift, prep each sub-test separately. Day 1 to 2: verbal comprehension with true/false/cannot-tell training at 15 seconds per passage-question pair. Day 3 to 4: numerical reasoning with chart and table reading at 15 seconds each. Day 5 to 6: diagrammatic reasoning with input/output rule discovery. Day 7 to 10: full 18-minute timed simulations that mirror the sub-test structure. Saville-specific: practice transitioning between sub-tests cleanly. Even a 15-second mental pause at sub-test boundary costs 1 question.
Shared prep: mental math fluency, vocabulary speed, and pattern recognition all help both tests. If you are uncertain which test is coming, drill foundational skills until invitation clarifies format. Then spend the final 3 to 5 days on test-specific pacing drills.
Order of prep if facing both: Cubiks first. Context-switching fluency is harder to build and transfers usefully to Saville (even though Saville does not require it, the broader pacing discipline transfers).
Which one you should actually prep for
Check the invitation vendor. Cubiks invitations come from cubiksonline.com or cubiksassessment.com. Saville invitations come from saville-assessment.com. This is the clearest single tell.
Industry is a secondary tell. UK energy, banking, and insurance lean Cubiks. UK public sector and accounting lean Saville. Both are used in UK professional services, so use vendor domain as the primary signal.
If you are applying broadly across UK graduate schemes without a confirmed test: Cubiks and SHL combined cover roughly 70 percent of the UK graduate cognitive testing market. Prep Cubiks as a secondary option behind SHL prep. Saville-specific prep is only worth it if you have a confirmed Saville invitation because the test is narrower in distribution.
Cubiks Logiks
Cubiks Logiks is a 12-minute general ability test covering numerical, verbal, and abstract reasoning in quick succession.
Saville Swift Aptitude
Saville Swift combines verbal, numerical, and abstract reasoning into a fast-paced aptitude battery. Common across UK public sector and professional services.
Related reading
Cubiks vs Saville FAQs
Same family, different pacing feel
Timed practice for both Cubiks Logiks and Saville Swift with the exact sub-test structure each requires.
Start Free Practice