What Is a Good Versant Score? Levels Explained 2026

🦈 Decoding a Versant score report is easy. Moving the number is the hard part — and it moves fastest when something corrects you while you speak. SpeakShark gives phoneme-level pronunciation feedback inside real conversation, and your AI teacher remembers the mistakes you keep making between sessions. Free tier, three sessions a day, no card. Start a free speaking session →

The Versant English Test reports an overall score from 20 to 80, and there is no official pass mark — Pearson sells Versant as a tool for employers to set their own cut score. The practical answer: roughly 58 and above lands at CEFR B2, the level most customer-facing English roles want, and roughly 47 to 57 lands at B1, which many non-voice roles accept. Everything below is traced to Pearson's own published documents, opened and quoted this week. Where the internet's numbers conflict — and on employer cutoffs they conflict badly — we show you the contradiction instead of picking the convenient one.

SpeakShark is a speaking-improvement tool, not an exam-prep provider, and is not affiliated with or endorsed by Pearson or any employer named on this page.

In this guide: the score scale at a glance · what the four subscores measure · what "Level 4" and "Level 5" mean · a good score for your job type · Versant to CEFR, GSE and IELTS · how we verified this · limits of the scale · raising a weak subscore · FAQ

The Versant score scale at a glance

Every figure in this table comes from Pearson's own Versant English Test product information sheet and its published sample score report.

Two things in that table do most of the work. First, the overall score is weighted, not averaged — so a very weak fluency subscore can drag your headline number below what the other three would suggest. Second, Pearson prints a CEFR level on the report, which is the closest thing to an objective definition of "good" that exists for this test.

If you have not sat the test yet, the task-by-task breakdown lives in our guide to Versant test questions, and the room-and-hardware checklist that quietly protects your pronunciation subscore is in Versant test setup: camera and headphones.

What the four Versant subscores actually measure

Pearson defines each subscore precisely. These definitions matter more than the overall number, because they tell you exactly what the machine is listening for.

• Sentence Mastery — "the ability to use basic and complex grammatical structures and sentence patterns accurately and appropriately." Scoring focuses on words, phrases and clauses in meaningful structures. In practice this is the Repeat and Sentence Builds tasks: can you hold a full sentence in working memory and reproduce it grammatically?
• Vocabulary — "the ability to understand, recall, and use common everyday words accurately and appropriately," judged at word, phrase, sentence and discourse level. Note the word everyday. Versant is not testing academic or technical range.
• Fluency — "the ability to maintain smooth speech flow and tempo," scored on "rate of speaking, the position and length of pauses, rhythm, and phrasing." This is the subscore that punishes hesitation, and the one candidates most often tank by pausing to plan a perfect sentence.
• Pronunciation — "the ability to articulate vowels and consonants in a native-like manner," with scoring focused on "accurate pronunciation and stress placement at the word and sentence level." Stress placement is explicitly in scope, which is why a technically correct sentence said with the wrong emphasis still loses points.

There is also the Intelligibility Index, which Pearson describes as measuring "various parts of a test taker's manner of speaking such as pronunciation, fluency, clarity of expression, and speed of speech." Pearson's own note is the useful one: "Even when spoken English proficiency is low, if intelligibility is high, it means that native speakers find the speech easy to understand."

The sample report shows why subscores beat the headline number as a study tool. That candidate scored 58 overall — but Fluency sat at Versant 51 (CEFR B1) while Pronunciation sat at 64 (CEFR B2+). Same person, a 13-point spread. The overall score tells you nothing about where to spend your practice time; the subscores tell you everything.

Versant test levels: what "Level 4" and "Level 5" actually mean

This is the most-searched and most-misanswered question about Versant, so here is the honest version.

The Versant English Test does not report a "Level 1-5" for your English. It reports a number from 20 to 80, four subscores, a GSE score and a CEFR level. If a recruiter tells you "you need Level 4," one of three things is going on:

1. They mean the Intelligibility Index. This is the only single-digit score on the report. Pearson's published sample shows Intelligibility 4, labelled "Good", with the explanation: "Listeners may require a little effort at times to understand some of the candidate's speech." That is the one level descriptor Pearson publishes openly. The full band table is not on Pearson's public pages, so any confident "Level 5 means X" you find on a forum is unofficial unless it came from your employer's own documentation.
2. They mean the CEFR band. The score report prints a ladder of <A1, A1, A2, A2+, B1, B1+, B2, B2+, C1, C2. Some companies rename these internally — "Level 3 = B1, Level 4 = B2" and so on — which is why the same "Level 4" means different things at different employers.
3. They mean an internal banding their HR team invented. Pearson's product sheet lists recruiting, promotion and placement as intended uses; setting the cut score is the customer's job, not Pearson's. Many BPOs bucket the 20-80 range into their own five tiers and never publish the boundaries.

The practical move is unglamorous: ask. "Is that the intelligibility number, the CEFR level, or your internal band — and what does it equal on the 20-80 scale?" A recruiter can answer that in one sentence. Guessing costs you weeks of misdirected practice. We cover how this plays out per employer in our guides to Accenture, Concentrix, Teleperformance, TCS and Cognizant, Deloitte USI, HCLTech and Amazon.

What is a good Versant score for your job type

Here is where most articles invent numbers. We are not going to.

No major employer publishes its Versant cut score. We checked, and the figures circulating online do not agree with each other. Prep sites variously claim that voice roles need "roughly 47+", that "50-54" is a common passing threshold, that "55-62" is typical for call centre roles, and that "most employers set a minimum between 56 and 69." Those claims span more than twenty points on an eighty-point scale, and none of them cites an official source. Reported as candidate folklore, they are informative. Reported as fact, they are misinformation.

What is defensible is anchoring to CEFR, because Pearson prints the CEFR level on the report and publishes the alignment. On that basis:

The Versant columns are our estimate, derived in the next section from Pearson's published anchor points — they are not a Pearson-published table, and we would rather say so than dress an inference up as an official figure. The CEFR column is the load-bearing part: if a job description says "professional working proficiency," B2 is what that phrase conventionally means, and Pearson's own numbers put the bottom of B2 at Versant 58.

One more honest note. A cut score is a business decision, and business decisions change with hiring demand. The same account can drop its threshold in a growth quarter and raise it six months later. Nobody online knows your employer's current number. Your recruiter does.

If you are preparing for a voice role specifically, the rehearsal that transfers best is not vocabulary lists — it is talking through real customer situations out loud, which is exactly the routine in our guide to English speaking practice for call center agents. Run your first free practice session →

Versant to CEFR, GSE and IELTS: the honest mapping

Pearson gives us two published pieces, and they connect cleanly.

Piece one — the anchor points. Pearson's sample score report states outright: "GSE 59/90 is equivalent to Versant 58/80." The same report shows four more paired values in its subscore breakdown:

Piece two — the GSE-to-CEFR table. Pearson's GSE alignment research paper (de Jong & Benigno, April 2017) publishes the official band boundaries: A1 = GSE 22-29, A2 = 30-42, B1 = 43-58, B2 = 59-75, C1 = 76-84, C2 = 85-90. The same appendix maps CEFR to IELTS: B1 = IELTS 5.5-6.5, B2 = 6.5-7.5, C1 = 7.5-8.5.

Chain those together and you get an approximate Versant-to-CEFR mapping. Fitting a straight line through Pearson's five published pairs gives Versant ≈ 0.65 × GSE + 19, which reproduces all five anchors within a point:

Read that table with three caveats. The B1 and B2 rows sit close to Pearson's published anchors and are the trustworthy ones. The C1 and C2 rows are extrapolated well past the last anchor point (GSE 69) and should be treated as rough. And the A1/A2 end is the least reliable of all, because the fit was built from mid-scale data.

There is also an open disagreement worth knowing about, because it affects any Versant-to-IELTS conversion you read. In that same paper, Pearson states that the IELTS-to-CEFR relationship published by the makers of IELTS "is understood to be overrated," and builds its table from ETS and Pearson research instead. So a Versant-to-IELTS conversion done via Pearson's chain will land roughly half a band lower than one done via IELTS's own CEFR claims. Both organisations are stating their position publicly; neither is hiding it; you should simply know which chain any given conversion used. If IELTS is your actual destination, start with our comparison of IELTS online vs paper.

How we verified this scale

A page about scores is only as good as its sources, so here is the full trail.

• Pearson's Versant English Test product information sheet (PDF, © 2020) — source for the 20-80 range, the four subscore definitions quoted verbatim, the Intelligibility Index, the ~15-minute / 63-question format, the six question types, the 0.97 correlation and split-half reliability figures, and the stated purposes.
• Pearson's published sample score report (PDF) — source for "a weighted combination of four diagnostic subscores," the GSE 59 = Versant 58 anchor, all five GSE/Versant/CEFR pairs, the CEFR ladder including the plus-bands, and the Intelligibility 4 = "Good" descriptor.
• Pearson's GSE alignment research paper (April 2017) — source for the GSE-to-CEFR boundaries and the CEFR-to-IELTS column, plus Pearson's stated view on IELTS's own CEFR claims.
• Pearson's current Versant test line-up for HR — source for the 2026 product range and durations, including the note that the older speaking-only test is retiring.
• Wikipedia's Versant entry — used only for uncontested history: the test began as Ordinate's PhonePass in the late 1990s, was renamed Versant in 2005, and Pearson acquired the division in January 2008.

What we refused to do: state an employer's cut score as fact. We found no employer publishing one. Where prep sites assert thresholds, we report the spread and label it unofficial, because a candidate who trains toward a fictional number wastes real weeks.

The limits of this scale and when a higher score will not help

Three honest limits.

1. The test you sit in 2026 may not be the 20-80 one. Pearson's current HR test line-up leads with the English Speaking and Listening Test (17 minutes), alongside a 4-Skills Essentials test (30 min), Writing (35 min), Professional English (60 min) and Placement (50 min) — and it flags the older speaking-only English Speaking Test as retiring soon. Pearson's current product pages emphasise reporting on the GSE 10-90 scale rather than the classic 20-80 Versant scale. So before you memorise a target number, confirm which Versant your employer administers and which scale your report will use. A "60" means very different things on a 20-80 scale and a 10-90 scale.

2. A score is a snapshot, not a diagnosis. Every test carries measurement error, and Pearson's own research literature is explicit that a reported score sits inside a range rather than on a point. Two attempts a week apart can differ by a few points without your English changing at all. Chasing a two-point gap is usually not a language problem.

3. Sometimes the score is not the bottleneck. If you freeze on live calls but score well on the test, more Versant prep will not fix your job. If you score badly because you rushed, panicked, or your headset clipped your consonants, that is a test-day problem, not a proficiency problem — see Versant test setup and how to prepare for the Versant test. And if the real issue is the fear of speaking at all, that is a separate, very trainable thing: our guide to overcoming the fear of speaking English covers the graded-exposure method.

When something else is genuinely better than a speaking app: if your target is an academic or immigration score, take an exam-prep course built for that exam — general speaking practice helps indirectly but is not exam coaching. If you need a signed proficiency certificate for a visa, only the official test issues one. And if your employer requires accent softening for a specific market, a dedicated accent-coaching product does that job more directly than conversation practice does.

How to raise the subscore that is holding you back

Because the overall score is weighted, the fastest gains come from lifting your weakest subscore, not from general study. Match the fix to the subscore Pearson names:

• Low Fluency — you are pausing to plan. The fix is volume of unscripted speech under mild time pressure, daily. Perfection is the enemy here: a fluent sentence with a small error scores better than a perfect sentence delivered after a three-second gap.
• Low Pronunciation — you need to know which sound failed, not that "pronunciation is weak." Feedback that names the phoneme is the whole game; feedback that just colours a word red tells you nothing actionable.
• Low Sentence Mastery — practise holding and rebuilding whole sentences aloud rather than trading single words. Repeat-and-extend drills map directly onto the Repeat and Sentence Builds tasks.
• Low Vocabulary — Versant tests everyday words, so read and speak ordinary situations, not academic word lists.
• Low Intelligibility with decent subscores — you are probably fast and flat. Slow down, group words into phrases, and put stress on the content words.

This is the specific job SpeakShark was built for: real conversation with real-time, phoneme-level pronunciation feedback while you speak — not isolated drills, and not word-level colour codes. Its differentiator matters more here than anywhere: cross-session memory. Your AI teacher remembers your recurring mistakes, your level and your topics between sessions, so the same dropped consonant gets tracked and targeted instead of re-flagged forever. That is what turns thirty sessions into compounding progress rather than thirty first days.

The free tier gives three sessions a day, forever, no card: five minutes each with teacher Sarah (American), 64 topics, basic feedback and 7-day history. Premium is $10/month or $69/year (about $5.75/month, 43% off) for unlimited 10-minute sessions with unlimited turns, four accent teachers (American, British, Australian, Canadian), 320 topics across 10 categories, advanced feedback and full history. Role Play mode is the one to use before a voice-role assessment — rehearse the customer call, not the vocabulary list. Interview coming up too? Our job-interview role-play scenarios cover that side. Not sure where you currently stand? Start by testing your English speaking level free.

🦈 Start your free SpeakShark session → — three conversation sessions a day, forever, no card. Phoneme-level feedback while you speak, and a teacher who remembers you tomorrow.

FAQ

Your score report is a map, not a verdict. Read the four subscores, find the one dragging the weighted total down, and spend the next three weeks on that one thing — out loud, daily, with something that tells you exactly which sound went wrong. Start your first free session on SpeakShark →

All Versant figures verified July 2026 from Pearson's own documents: the Versant English Test product information sheet, the sample score report, the GSE alignment research paper (April 2017) and the current Versant test line-up. Versant-to-CEFR score estimates are our own derivation from Pearson's published anchor points and are clearly labelled as estimates, not official figures. Employer cut scores are set by employers, change over time, and are not published — confirm yours with your recruiter. SpeakShark is a speaking-improvement tool, not an exam-prep provider, and is not affiliated with or endorsed by Pearson, IDP, the British Council or any employer named. No affiliate links.