SVAR Test 2026: Complete Guide to Pattern, Prep, and Score

SVAR is a 16 to 20 minute, AI-graded spoken-English test that companies use to screen voice-process and customer-experience candidates. It sits between your online aptitude round and your HR interview. Clear it, and the rest of the process opens up.

This guide covers the full picture: what SVAR actually tests, how the AI grades your audio, where SVAR fits alongside Versant and Accenture’s own communication test, what score you need by role type, and a prep plan built around how the AI actually works.

What is the SVAR test — and who uses it in 2026

SVAR stands for Spoken English Voice and Accent Recognition test. It’s an optional module within the wider AMCAT (Aspiring Minds Computer Adaptive Test) platform, now run by SHL India after its 2019 acquisition of Aspiring Minds. SHL reports that AMCAT is accepted by over 3,000 employers in India; SVAR specifically is used by employers where spoken-English fluency is a job requirement.

The test runs 16 to 20 minutes, browser-based, with a webcam and microphone. The older telephone-based delivery has mostly been retired. No human listens to your audio in the standard scoring path.

Three employer types consistently include SVAR in their hiring pipeline.

Voice-process and BPO firms use it as a primary English fluency screen before any technical or HR round. Concentrix, HGS, and Tech Mahindra BPS are among the consistent users.

Mid-size IT services firms with international clients use SVAR for voice-adjacent roles where written English alone doesn’t cover the job: client servicing, L1 support, and pre-sales communication.

B2B SaaS and customer-success teams have adopted SVAR or an equivalent in 2025 and 2026 for roles where English communication happens in client calls, demo recordings, and async video updates.

For a detailed section-by-section breakdown with full tips per section, the AMCAT SVAR test pattern and preparation guide covers each of the six sections in depth. This guide zooms out to the strategy layer: how to approach SVAR as part of the broader communication-test picture.

The six sections at a glance

The standard SVAR test has six sections and around 45 questions. All six sections test spoken production or listening comprehension; none are written. Here’s the structure:

Section	Approximate questions	What it measures
Reading	12	Pronunciation accuracy, fluency, pacing
Listening (repeat)	10	Listening accuracy, verbatim repetition
Listening Deductions	6–8	Comprehension of spoken workplace dialogue
Grammar (audio)	8	Identifying grammatically correct spoken sentences
Error Identification	6–8	Spotting error type in a heard sentence
Speaking (open)	3–5	Extended spoken production on a prompt

The table above gives structural planning numbers. Exact counts vary slightly by test version.

Sample question formats, one per section:

• Reading: A sentence appears on screen — for example, “The project deadline has been moved to the last Friday of this month.” Read it aloud after the beep. The AI scores pronunciation, pacing, and full-word articulation.
• Listening Repeat: You hear: “Please submit all client documents before five thirty in the evening.” Repeat it exactly as heard within the response window. Paraphrasing counts against you.
• Listening Deductions: A 60 to 90 second workplace conversation plays. You answer two or three multiple-choice questions about the factual content. The AI grades MCQ accuracy, not your interpretation.
• Grammar: You hear three spoken sentence variants — for example, “A: she don’t like coffee / B: she doesn’t likes coffee / C: she doesn’t like coffee” — and select the correct one.
• Error Identification: A sentence with a deliberate error is read aloud. You identify the error type from a list: subject-verb agreement, preposition, tense, or article.
• Speaking: You see a prompt — for example, “Describe a time you solved a problem at work or in a project.” Prep window: 30 to 45 seconds. Response window: 45 seconds. Structured is better than impressive.

How the AI grader actually scores your audio

The grading pipeline runs in five stages. Knowing each one changes how you prepare.

Stage 1: Automatic Speech Recognition

Your audio is transcribed to text. The ASR model is trained on Indian-accented English, which makes it reasonably forgiving of pronunciation variation. What it punishes is mumbling and incomplete words, because those don’t transcribe cleanly. Every scoring stage that follows depends on a legible transcript.

Stage 2: Phonetic similarity

Your transcription is aligned at the phoneme level against reference pronunciations. The system computes a Phoneme Error Rate, which measures how far your phoneme sequence deviates from the expected one. A January 2026 peer-reviewed study in Frontiers in Communication documents the ECAPA-TDNN plus Whisper pipeline used for Indian-learner accent assessment, with reported classification accuracy in the 88 to 93% range across proficiency levels. SVAR and Pearson’s Versant use closely related pipelines.

Stage 3: Pace and fluency

The system measures words-per-minute, silence ratios, and filler-word frequency. Conversational speech (approximately 110 to 130 WPM) scores higher than very slow or very fast delivery. This is where automaticity matters. You don’t need a wide vocabulary; you need the vocabulary you have to come out automatically, without pauses to search for words.

Stage 4: Content match

For Reading and Listening Repeat sections, the system checks how close your transcription is to the source text. For Speaking sections, it checks for topic-relevant keywords and structural coherence. No section grades your opinion or the quality of your argument.

Stage 5: Composite scoring

The section sub-scores are weighted together into a percentile on a 0 to 900 scale.

Two prep implications follow directly from this pipeline.

First, the AI does not hear your accent the way a human does. It measures consistency and phonetic accuracy at the phoneme level. An Indian student speaking their natural accent clearly and confidently will outscore the same student attempting a half-formed American accent. Accent mimicry introduces inconsistency, which is the exact thing the system penalises.

Second, mumbling hurts you twice: once at the ASR stage, because the model can’t transcribe what it can’t parse, and again at the fluency stage, because incomplete words and long silence gaps read as disfluency. Speaking clearly at conversational pace, even with minor grammatical slips, beats speaking quietly and carefully.

SVAR vs Versant vs Accenture: which test you’ll actually face

Three AI-graded spoken-English tests dominate Indian campus hiring in 2026. They measure roughly the same underlying ability. The section formats, scoring scales, and the employers who require each one differ.

	AMCAT SVAR	Versant (Pearson)	Accenture Communication Assessment
Provider	SHL India	Pearson	Accenture (proprietary)
Sections	6	6 task types	6
Questions	~45	Varies by task	63
Duration	16–20 min	~15 min	25–35 min
Scoring scale	0–900 (percentile)	20–80	AI-graded (recruiter cutoff)
Primary employers	Voice BPOs, customer-success firms	Deloitte USI, consulting firms	Accenture (ASE and Advanced ASE tracks)
Prep transfers?	—	Yes, to/from SVAR	Partially (format differs more)

Accenture uses a proprietary format with 6 sections and 63 questions. The Accenture Communication Assessment guide covers the full pattern, including question-type examples and the 20 to 30 second response window.

If you’re applying to Deloitte or another firm that uses Versant, the Deloitte Versant test guide covers the six task types and the 20 to 80 scoring scale in detail.

One important note on prep portability: the skills that clear SVAR transfer well to Versant, because both tests reward the same behaviour: clear, consistent, paced spoken English. The Accenture test has a different structure. Its Question-and-Answer and Jumbled Sentences sections don’t have direct SVAR equivalents, but the underlying fluency and listening accuracy skills carry over cleanly.

What score you need by role type

AMCAT scores are reported as percentiles per section on a 0 to 900 scale. Companies don’t publish hard SVAR cutoffs publicly. The table below is a planning benchmark, not a guarantee:

Role type	Typical percentile target
Domestic voice process (inbound/outbound BPO)	50th percentile and above
International voice process	75th percentile and above
Voice-adjacent IT services (client servicing, L1 support)	60th to 70th percentile range
Premium customer-success (B2B SaaS, global clients)	75th percentile and above

Your score report includes a recruiter-facing percentile breakdown across sections. The number recruiters see is comparative against the testing pool, not absolute, so the cutoff bar can shift across test cohorts.

One practical implication: if you score below your target in a specific section, focus your retake prep on that section rather than retaking the full test hoping the composite average improves.

How to prepare — a practical plan

The core insight from the AI grading pipeline is this: SVAR doesn’t test vocabulary range or grammar perfection. It tests whether you can produce clear, fluent English at conversational pace under timed conditions. That distinction changes what good preparation looks like.

What works

Read English text aloud every day, at conversational speed, on the same setup you’ll use for the test. A quiet room, a working microphone, and 20 minutes per day outperform intensive cramming sessions the night before.

Diagnose first. Record yourself reading two or three paragraphs from any English newspaper editorial and listen back. Look for three specific patterns: mumbling at the end of sentences, over-long pauses between phrases, and words you skip when reading quickly (usually “the,” “a,” and connector words like “which” and “that”). Fix the patterns you find, not the ones you assume you have.

Grammar review should focus on Indian-English error patterns specifically: subject-verb agreement (“she don’t” versus “she doesn’t”), preposition collocations (“discuss about” is wrong; “discuss” alone is correct; “return back” is redundant; “explain to me” not “explain me”), tense consistency, and article use. These are the recurring patterns in the Grammar and Error Identification sections.

For the Speaking section, structure beats content every time. A three-point frame (“First… Second… Finally…”) delivered at conversational pace consistently outscores an unstructured, more nuanced response. Use the 30 to 45 second prep window: write down your three points on the provided rough sheet rather than just thinking about them.

What doesn’t work

Memorising vocabulary lists wastes preparation time. SVAR doesn’t reward rare vocabulary. The scoring system rewards automaticity with existing vocabulary.

Accent training in the wrong direction hurts. Practising to sound American or British introduces hesitation and phonetic inconsistency. Speak in your natural accent with clarity.

For a more detailed section-by-section daily schedule, the AMCAT SVAR pattern and tips guide includes a full 7-day plan with specific practice formats per section. The broader communication skills required across GD rounds, HR interviews, and written assessments are covered in the communication skills for campus placements guide.

Day-of-test setup

Run a full system check 30 minutes before your slot: webcam on, microphone input level correct, browser (Chrome or Edge) updated, background noise minimal. Close any tabs playing audio or video. The test doesn’t have a pause function once started.

Technical failures during the test are routed to SHL India’s support team; your placement cell coordinator should have the escalation contact. If you lose connection mid-section, note the time and contact support immediately; don’t restart the test without guidance.

After the test, your score report is typically available within 24 hours. The report shows your section-wise percentile breakdown and a composite score. Share it with your placement cell or the recruiting company as instructed in the drive communication; don’t assume they receive it automatically.

If you’re preparing for voice-process or customer-experience roles at multiple companies and need to understand how the spoken-English dimension fits into the broader placement communication picture, the communication skills guide covers GD rounds, HR interviews, and written verbal tests alongside SVAR in one coherent framework.