OPIc Levels and Grades: The Full Scale Explained

The OPIc reports a grade rather than a score, and that grade is a band on the ACTFL proficiency scale. In Korea the reported scale runs from Novice Low up to Advanced Low across nine grades, with Intermediate Mid split into three. This guide is the reference version: every band, what the official wording actually says, and which levels map to which realistic uses.

If you want to practise the unscripted speaking that every one of these bands is built to measure, SpeakShark is free with three AI conversation sessions a day and no card.

In this guide: the full ladder · what Low, Mid and High mean · Novice and Intermediate Low · Intermediate Mid · Intermediate High · Advanced Low · the IM1, IM2 and IM3 split · where the scale stops being precise · bands and jobs · which guidelines apply · what it does not convert into · how we researched this guide · FAQ

Key takeaways

• The Korean OPIc reports nine grades, Novice Low to Advanced Low. The live Oral Proficiency Interview reports Novice Low to Superior and does not subdivide Intermediate Mid.
• opic.or.kr publishes exactly seven named band descriptors, and a single one of them covers IM1, IM2 and IM3 together. No separate IM1, IM2 or IM3 wording exists on the official site.
• ACTFL's reliability study found two certified raters assigned the identical band on the English OPIc 59.6 percent of the time, and landed within one band 94 percent of the time.
• ACTFL defines a sublevel as the quality of performance at the level plus proximity to the next major level, not as a separate measurement.
• ACTFL's published workplace minimums put a receptionist or cashier at Intermediate Mid, a tour guide or firefighter at Intermediate High, and a K-12 language teacher or police officer at Advanced Low.

The full grade ladder in one table

Two scales are in play. The Korean OPIc reports part of the ACTFL scale, and the rest of the ACTFL scale exists above it but never appears on a Korean result. The official Korean site puts the difference in one line: the OPIc covers Novice Low to Advanced Low and subdivides the Intermediate Mid band, while the OPI covers Novice Low to Superior and does not subdivide it.

Two things in that table are worth stopping on. Three consecutive rows carry the same descriptor, which is the single most important fact about the Korean scale. And the scale does not end at Advanced Low, it stops being reported there.

If your question is simply whether Advanced Low is the maximum, our Advanced Low guide answers it in full, including which forms can and cannot award it.

What Low, Mid and High actually mean

Every ACTFL grade at the Novice, Intermediate and Advanced levels is a major level plus a sublevel. The structure is fixed: those three major levels are each divided into High, Mid and Low, while Superior and the Distinguished level above it are not divided at all. The Familiarization Guide describes the scale as a continuum with five major levels, Novice, Intermediate, Advanced, Superior and Distinguished.

The sublevels are defined once and apply identically at every major level. ACTFL's Familiarization Guide puts it like this:

Low: Baseline performance at the level. … They demonstrate little ability to carry out communication functions that typify the next-higher major level.

Mid: Solid performance at the level. Individuals at this sublevel demonstrate skill in carrying out the functions of the major level in all level-relevant communication situations. They may also exhibit partial ability to carry out communication tasks that typify the next-higher major level.

High: Performance with quality and quantity at the level. Individuals at this sublevel communicate with ease and confidence when performing the functions of the level. They are capable of functioning much of the time at the next higher major level, but they are unable to sustain language at that level.

Read that once and the whole ladder becomes readable. Novice High is not a slightly better Novice Mid in some vague way, it is Novice performance with quality and quantity, from a speaker who can already function much of the time as an Intermediate but cannot hold it there. Intermediate Low is not a weak Intermediate in general, it is baseline Intermediate with almost nothing of Advanced showing yet.

ACTFL states the logic of the split even more directly elsewhere. A rating at any major level confirms sustained performance across all of the criteria of that level, and the sublevel is determined by the quality of the performance at that level and the proximity to the next higher major level. The major level is the measurement. The sublevel is a judgment about where you sit inside it.

That single sentence explains more about your grade than any band chart, and it is the reason the rest of this guide keeps returning to major levels rather than notches.

What the Novice bands and Intermediate Low cover

The Korean site publishes a named descriptor for each of Novice Low, Novice Mid, Novice High and Intermediate Low, and those are the versions to read if you hold one of those grades. We are not going to paraphrase wording we cannot quote from a primary source, so the honest summary is the structural one.

Applying ACTFL's own sublevel rule gives you the shape of each band:

• Novice Low is baseline Novice, with little ability to do what the Intermediate level typifies.
• Novice Mid is solid Novice across all level relevant situations, sometimes with partial Intermediate ability appearing.
• Novice High is Novice with quality and quantity, sitting closest to the Intermediate boundary.
• Intermediate Low is baseline Intermediate: the Intermediate functions are there, but Advanced features are largely absent.

There is one thing ACTFL says about the Novice range that no prep site repeats, and it is blunt. In the same document where it lists recommended proficiency levels for real jobs, ACTFL states that Novice level proficiency is not included in those recommendations, because speakers are not able to demonstrate enough functional ability for the workplace at this level.

That is not a comment on effort or intelligence. It is a statement that the Novice bands describe a speaker who cannot yet hold a workplace exchange, which is why no job appears against them. If you are sitting in that range, the useful next step is volume of unscripted talking rather than test tactics. Practising English speaking alone and free methods to raise fluency are the relevant starting points, and why you can read English but cannot speak it covers the specific gap most people in this range are actually facing.

What Intermediate Mid actually says

This is the band the Korean scale bothered to subdivide, and the one the published recruitment notices we have seen reference most visibly, so it is worth reading in both versions. Nobody publishes a distribution, so treat that as an observation about the notices rather than a count.

The official Korean descriptor for Intermediate Mid describes a speaker who can string sentences together and speak naturally not only on everyday topics but in personally familiar situations, who experiments with a range of sentence forms and vocabulary, and who can hold a long conversation provided the other person shows a little consideration.

ACTFL's current Proficiency Guidelines describe the same band from the other direction:

Speakers at the Intermediate Mid sublevel are able to handle successfully a variety of uncomplicated communication tasks in straightforward social and transactional situations. … Their speech largely consists of strings of sentences; it may contain pauses, reformulations, and self-corrections as they search for adequate vocabulary and appropriate language forms to express themselves

The load-bearing phrase is strings of sentences. Intermediate Mid is a sentence level band. Pauses, reformulations and self-corrections are written into the definition rather than counted against you, which surprises most people preparing for it.

ACTFL's Examinee Handbook adds the part that matters for the Korean split. A worked Intermediate Mid speaker creates sentence level responses consistently throughout the test, asks questions, provides information to handle a social transaction, does so with good quality and an abundance of language, and can generally demonstrate some features of the Advanced level. So Advanced features already appear inside a clean Intermediate Mid performance. They are simply not sustained.

For what employers do with this band, and the practical distance between the three IM grades, see our IM2 guide, which covers the requirement side in detail.

What Intermediate High actually says

Intermediate High is unusual among proficiency descriptors because failure is written into the definition rather than left implicit.

They often can narrate and describe in all major time frames using connected discourse of paragraph length. When they do so, their speech exhibits one or more features of breakdown, such as inability to fully carry out narration or description in the appropriate major time frame, difficulty maintaining paragraph-length discourse, or a reduction in breadth and appropriateness of vocabulary.

Read that carefully. An Intermediate High speaker does reach Advanced level behaviour. They narrate across past, present and future, and they build paragraph length stretches. The band is defined by what happens next: the tense slips, or the paragraph collapses back into sentences, or the vocabulary narrows under the load.

That is why Intermediate High is the most useful diagnostic band on the scale. If you can already produce Advanced shaped speech but it comes apart when the topic gets hard, you are describing the band, not falling short of it.

Our Intermediate High guide covers who asks for IH and how to close the gap to it. For the underlying habit, twelve daily speaking drills from B1 to C1 and techniques to stop hesitating both target the breakdown features named above.

What Advanced Low actually says

Advanced Low is where paragraph length connected discourse stops being occasional and becomes the requirement.

At the Advanced Low sublevel, speakers demonstrate the ability to narrate and describe in the major time frames of past, present, and future with some control of aspect, although their narrations and descriptions tend to be handled separately rather than interwoven. Speakers at the Advanced Low sublevel combine and link sentences into connected discourse of paragraph length.

The official Korean descriptor lands in the same place from a teaching angle. It describes a speaker who manages verb tense consistently when narrating events, uses varied adjectives when describing people and objects, places conjunctions appropriately so the binding between sentences is strong, can skilfully construct paragraph structure, and can explain and resolve problems even in unfamiliar and complicated situations.

ACTFL's Examinee Handbook is refreshingly honest about what Advanced Low sounds like in practice. It describes a speaker able to just hold onto the Advanced level by producing short paragraphs and minimally controlling the corresponding linguistic functions in the appropriate major time frames, where the listener may need to make an effort to understand the intent of the message.

That is the band as its own examiner describes it: holding on, minimally, with the listener working. Anyone telling you Advanced Low means polished English is describing something else. Our Advanced Low guide covers how to get there and which form to sit.

Where the IM1, IM2 and IM3 split comes from

Here is what is officially established, and it is shorter than you expect.

The Korean site states that the OPIc subdivides the Intermediate Mid grade, and it states the ordering: IM3 above IM2 above IM1. That is the entire published content of the split. The same page carries seven named band descriptors, for AL, IH, IM, IL, NH, NM and NL, and exactly one of them covers Intermediate Mid. There is no IM1 descriptor, no IM2 descriptor and no IM3 descriptor anywhere on it.

The gap widens when you look upstream. No ACTFL or Language Testing International document mentions IM1, IM2 or IM3 at all. The strings do not appear in the OPIc Examinee Handbook, the Familiarization Guide or either ACE report. The suffix is a Korean reporting layer construct, and the rating machinery above it does not know the numbers exist.

The closest thing to an official explanation of why Intermediate Mid would need splitting is a general statement about the Mid sublevel, not a rule for cutting it. ACTFL's Examinee Handbook says there are many kinds of speakers that match the description of the Mid sublevel, depending on the amount of speech produced, the quality of the speech in all of the functions associated with the level, and the degree to which the speaker controls language features from the next higher major level.

Three dimensions of variation inside one band. That tells you why Mid is broad. It does not tell you where two cuts go.

On the descriptors you will find elsewhere. Sets of per level wordings circulate widely, usually along the lines of IM1 answering at paragraph level on various topics, IM2 giving concrete explanation and description with natural flow, and IM3 answering logically even on complex topics. None of them are traceable to opic.or.kr, ACTFL or Language Testing International, so confirm them with the exam body before relying on them.

Be especially careful with the IM1 version. Paragraph level discourse is the Advanced level threshold in ACTFL's own descriptors, which makes it an odd thing to attribute to the lowest Intermediate Mid grade.

On why Korea split the band at all. The usual explanation is that Korean test takers cluster there and domestic employers wanted finer resolution. That is plausible and completely undocumented. The official site states that the split exists and never states why. Treat it as background rather than fact, and check anything decision critical with the exam body.

On grade distribution figures. Numbers like "IM2 is the most common grade" or specific percentages per band are not published by any Korean primary source we could find. The only official OPIc distribution available covers ACTFL's international test population, and it looks nothing like the Korean picture, so it cannot be used to describe Korea. More on that below.

If your requirement is stated as IM2, the practical guidance is in our IM2 guide rather than here, since this page is the reference and that one is the strategy.

Where the scale stops being precise

An Official ACTFL OPIc is rated twice. Official here is a product tier rather than a synonym for the real exam: Language Testing International states that commercial OPIcs are rated by a single certified rater, while an Official or Certified test gets two. On that tier, two ACTFL certified raters score the sample anonymously, meaning the second does not see the first rating, and the two ratings must agree exactly on both the major level and the sublevel to be final. When they do not, the test goes to a third rater for anonymous arbitration.

Now the number almost nobody quotes. ACTFL's own statistical analysis reports that the first two raters assigned exactly the same band on the English OPIc 59.6 percent of the time. The share where the first two ratings were either exact or adjacent was 94 percent.

Set that beside ACTFL's own definition of what a sublevel is, quality of performance plus proximity to the next major level, and the conclusion writes itself:

1. The major level is the hard signal. Novice, Intermediate or Advanced is what two trained raters reproduce reliably.
2. One notch is roughly the width of the measurement. Exact agreement sits near 60 percent while exact plus adjacent agreement sits at 94 percent, which is what a boundary judgment looks like in data. Adjacent agreement on its own accounts for the remaining 34 percent.
3. The arbitration process exists precisely because of this. The third rater is not a rare failsafe, it is a designed response to disagreement at the notch.
4. Korea then split Intermediate Mid into three and published one descriptor for all three and no criteria for the cuts.

The practical reframe: stop optimising for the notch and optimise for the major level jump, because that is the part of your grade that is reproducible. Chasing IM2 to IM3 as if it were a fixed target is chasing a distinction ACTFL raters never make, since they assign Intermediate Mid as one undivided band and disagree on the exact band roughly four times in ten even at that resolution.

Two more precision limits belong here. ACTFL concedes that a badly chosen self assessment can produce a rating that does not reflect genuine proficiency, and that a rater who suspects this makes a standard notation at evaluation time which can be conveyed to the examinee or client. Separately, Forms 4 and 5 do not floor out at their lowest listed rating, they return no rating at all below it. Our self assessment guide covers the five form table and that mechanic in full, so we will not repeat it here.

One last caution about sourcing. ACTFL's own ACE Part A report lists the possible assignable ratings and omits Intermediate Mid from the list, even though the form table in the same document plainly includes it. That is an error in an official document, not evidence about the scale. It is a useful reminder that single sentences, including single sentences on prep sites, are not a sound basis for a decision about your career.

Which bands map to which jobs

The only primary source that maps proficiency bands to real occupations is ACTFL's own list of recommended minimum levels. It is worth reading as published, because it is far more grounded than the job charts circulating on prep sites.

Two caveats before you use this. It is ACTFL's international recommendation about language demand in a role, not a hiring rule and not a Korean employer policy. And it is stated as examples rather than as an exhaustive mapping.

Even so, it reframes the question people actually ask. Intermediate Mid is the level at which ACTFL considers language workable in a customer facing job at all. Intermediate High is where roles carrying responsibility under pressure begin. Advanced Low is where ACTFL puts teaching and policing, which are jobs where being misunderstood has consequences.

For what Korean employers specifically ask for, our IM2 guide and Intermediate High guide cover the published requirement patterns. If your process involves a recorded interview rather than a certificate, HireVue interview questions and English job interview role plays are the more relevant preparation.

There is one distribution figure worth knowing, with a warning attached. ACTFL published a band by band breakdown of English OPIc ratings in its international test population, and about 62 percent of those takers were classified at the Advanced level, with Intermediate Mid at 4.40 percent. That is the Language Testing International population, not the Korean one, and using it to reason about Korean grades would be badly wrong. It is here only so you recognise the numbers if you meet them.

Which version of the guidelines your grade is rated against

This is a genuine open question and no source resolves it.

ACTFL's 2024 Familiarization Guide says ratings are assigned against the ACTFL Proficiency Guidelines 2024 for Speaking. The ACE Part A report, which carries publication number AAR/OPIc/ACE/I-2020-002 on its cover and cites nothing later than 2015, and the 2018 Examinee Handbook both point at the 2012 guidelines. The 2023 label attached to that ACE report in most citations, including our own source list below, comes from its filename rather than from anything printed inside it. The Korean site links out to the 2024 guidelines while displaying its own Korean language descriptors that predate them.

The wording was substantively revised between the two versions, so this is not a cosmetic difference. Yet no document we could find states which version governs a Korean rating today. If your decision depends on it, ask the exam body directly rather than trusting any guide, this one included.

The reassuring part is that the disagreement is at the wording level, not the structural one. Both versions use the same major levels, the same three sublevels, and the same fixed rule that Superior and Distinguished carry no sublevels.

What the scale does not convert into

The most common request after "what are the levels" is a conversion table, usually to CEFR or to another speaking test. Here is what is actually published.

Language Testing International states that the OPIc rates the full range of the ACTFL scale from Novice through to Superior, and that it assesses Levels 0 to 3 on the ILR scale and A1 to C2 on CEFR. Those are ranges, describing what the instrument can report against. They are not band level equivalences, and no band by band CEFR or ILR conversion table is published by ACTFL, Language Testing International or opic.or.kr that we could retrieve. Every clean conversion chart in circulation comes from prep sites rather than from an exam body.

The same applies to comparisons with other tests. If you are choosing between exams rather than converting between them, OPIc vs TOEIC Speaking covers the trade off honestly, and the TOEIC Speaking guide covers that route on its own terms.

On format, a short note is enough here. The Korean operator publishes a test of 12 or 15 questions, two discrete counts rather than a range, differing by the Self Assessment step you choose before the test starts. It gives 40 minutes for the questions inside a session of roughly 60 minutes including orientation. The same site's process diagram separately shows roughly 7 questions in a first session and roughly 5 to 8 in a second, with a difficulty re-selection step between them, which is the site's own approximation rather than a published count. ACTFL publishes no question count at all in the documents we read.

The two exam bodies also describe the length differently. The ACTFL Familiarization Guide says the test takes 20 to 40 minutes on average, while the ACE report describes a 20 to 40 minute recorded speech sample, which would make that figure speaking time rather than seat time. We could not reconcile the two, so plan around the Korean operator's 60 minute session and treat the shorter numbers as unsettled. Everything about what you will actually be asked lives in our OPIc question guide.

If you have no idea which band you are near, guessing from a table is the wrong method. Test your English speaking level free is a faster read on where you stand, and thinking in English plus role play scenarios are the two habits that move a major level rather than a notch. Daily unscripted talking on SpeakShark's free tier builds the sustained speech a major level is defined by, which is the part no chart can give you.

How we researched this guide

The scale structure, the Intermediate Mid subdivision, the ordering of IM3 above IM2 above IM1, the seven named band descriptors, and the 12 or 15 question format all come from the official OPIc Korea site, read directly rather than through summaries. The Intermediate Mid and Advanced Low Korean descriptors are rendered here from that page.

The sublevel definitions and the two rater process come from ACTFL's OPIc Familiarization Guide. The statement that the sublevel is quality plus proximity, the occupation minimums, the not ratable rule for Forms 4 and 5, and the self assessment mismatch note come from ACTFL's ACE Report Part A. The 59.6 percent exact and 94 percent adjacent rater agreement figures and the population distribution come from ACE Report Part B.

The Mid sublevel variation statement and the Intermediate Mid and Advanced Low worked examples come from the ACTFL OPIc Examinee Handbook. The verbatim Intermediate Mid, Intermediate High and Advanced Low wordings come from the ACTFL Proficiency Guidelines 2024 for Speaking, with the fixed sublevel structure confirmed against the 2012 edition. The CEFR and ILR ranges come from Language Testing International.

What we could not verify is stated as such in the text rather than smoothed over. There are no published descriptors for IM1, IM2 or IM3 individually, and no published criterion, threshold or rater instruction for where the two cuts fall. There is no official statement of why Korea subdivided the band. There is no Korean grade distribution from a primary source, and ACTFL's international distribution cannot substitute for one.

There is also no band level CEFR or ILR conversion table from any exam body, and the official documents disagree about whether the 2012 or 2024 guidelines govern a rating today. ACTFL publishes no question count anywhere we looked, so the 12 or 15 figure rests on the Korean site alone, and its block structure of roughly 7 questions then roughly 5 to 8 is a process diagram rather than a stated count. The two bodies also disagree about length, since the Familiarization Guide's 20 to 40 minutes reads as seat time while the ACE report's 20 to 40 minutes reads as recorded speech, and nothing we found reconciles them. We could not date the ACE Part A report from its own contents beyond publication number AAR/OPIc/ACE/I-2020-002, so the 2023 in its filename should not be read as a publication year.

We flagged one further problem: an outright error in ACTFL's own ACE Part A report, where the list of assignable ratings omits Intermediate Mid while the form table in the same document includes it.

Practise speaking, from SpeakShark

SpeakShark is an AI English speaking practice app, and it is our pick for anyone trying to move up the OPIc ladder rather than just decode it. Every band on this page is defined by what you can sustain in unrehearsed speech: strings of sentences at Intermediate Mid, paragraph length discourse that holds at Advanced Low. That is built by producing a lot of speech against something that answers back unpredictably. You talk, the AI answers, and you get feedback on pronunciation, grammar and fluency while the conversation is still going.

Start free with three sessions a day and no card. Limits are on the pricing page, and how it works walks through a full session.

We are a speaking improvement tool. We are not an exam preparation provider and we are not affiliated with ACTFL, Language Testing International, ETS or any exam board. For test format, rules, booking and official practice material, go to the exam body directly. Use SpeakShark to make your spoken English stronger, and use official material to learn the test.

• Site: speakshark.com
• Support: speaksharksupport@gmail.com
• Free tier: 3 AI conversation sessions per day, no card

Sources

• OPIc Korea, about the test: the nine grade Korean scale, the OPIc versus OPI comparison, the Intermediate Mid subdivision and the IM3 above IM2 above IM1 ordering, the seven named band descriptors, the 12 or 15 question format and the process diagram behind it
• ACTFL OPIc Familiarization Guide: the Low, Mid and High sublevel definitions in full, the five major level continuum from Novice to Distinguished, the average length of 20 to 40 minutes, the two rater process and third rater arbitration
• ACTFL ACE Report Part A, general test information, publication number AAR/OPIc/ACE/I-2020-002 despite the 2023 in the filename: the sublevel as quality plus proximity, the recommended occupation minimums, the not ratable rule for Forms 4 and 5, the self assessment mismatch notation
• ACTFL ACE Report Part B, statistical analysis: 59.6 percent exact rater agreement, 34.1 percent adjacent agreement and 93.7 percent exact plus adjacent on the English OPIc, the international band distribution
• ACTFL OPIc Examinee Handbook: the three dimensions of variation inside the Mid sublevel, the Intermediate Mid and Advanced Low worked examples
• ACTFL Proficiency Guidelines 2024 for Speaking: the verbatim Intermediate Mid, Intermediate High and Advanced Low descriptors
• ACTFL Proficiency Guidelines 2012 for Speaking: confirmation that Advanced, Intermediate and Novice each divide into High, Mid and Low while Superior does not
• ACTFL Proficiency Guidelines landing page: the current published editions of the guidelines
• ACTFL, Oral Proficiency Interview by Computer: the assessment overview
• Language Testing International, about the OPIc: the full ACTFL range statement and the ILR 0 to 3 and CEFR A1 to C2 reporting ranges
• ACTFL OPIc Examinee Handbook, mirrored copy: secondary access to the handbook used for cross-checking

FAQ