CELPIP Speaking Samples With Scores: Band 7 vs 9 Compared

Everyone wants speaking samples. Here's what most sample pages actually give you: one transcript, no score, no explanation, no way to tell what band it's at — so you read it, nod, and learn nothing about your own answers. This post does the opposite. Three task types, each with two full sample answers on the same prompt: a band 7 and a band 9. Then a criterion-by-criterion verdict using the actual four criteria raters score — content/coherence, vocabulary, listenability, task fulfillment. By the end you'll know the gap precisely enough to hear it in your own recordings.

One honest note first: these samples were written for this post as teaching examples — they show the levels, but they're not real student recordings. The verdicts are how our graders would score them. Want a real baseline? Record yourself and take a free mock test → celpipuni.com/free-mock-test.

TL;DR

The difference between band 7 and band 9 CELPIP speaking is not vocabulary. Across every sample pair below, the 9s use ordinary words — the gap is interpretation, specificity, rhythm, and register. Each pair shows the same prompt answered both ways, with a four-criteria verdict.

Sample pair 1: advice (Task 1 style)

Prompt: your friend wants to get fit but has never exercised and doesn't know where to start. Advise him.

Band 7:

"I think you should start exercising. First, you can join a gym. Gyms have many machines and trainers who can help beginners. Second, you should wake up early and go for a run in the morning. Morning exercise is very healthy and it will give you energy for the day. Third, you can find a friend to exercise with you, because it is easier to continue when you have a partner. If you do these things, you will become fit. It is important to be patient and not give up. I believe you can do it."

Band 9:

"So, here's what I'd do — and I'd start smaller than you think. Don't join a gym. Not yet. January gyms are full of people who pay for three months and show up four times, and you'd be one of them by February. Instead: walk. Twenty minutes after dinner, every day, this week. It's boring, that's the point — boring is what survives.

Once that's a habit — give it three weeks — add one thing you'd almost enjoy. Swimming, badminton, a beginner class where everyone else is also embarrassed. You're not building fitness yet, you're building the identity of someone who moves.

And track it somewhere you'll see it — a calendar on the fridge, an X through each day. You're not 'a fitness person' yet. Give it a month."

Sample pair 2: describing a scene (Task 3 style)

Prompt: describe a park scene — a man reading on a bench, children playing, an elderly couple walking a small dog.

Band 7:

"This picture shows a park. There is a man sitting on a bench. He is reading a book. Behind him, there are some trees. On the grass, two children are playing with a ball. They look happy. In the background, there is an elderly couple. They are walking with a small dog. The dog is white. The weather looks sunny and warm. The park seems like a nice and peaceful place."

Band 9:

"Late afternoon in a neighborhood park — the light's going long and gold, which is why I'd say it's a weekend rather than a workday. On the nearest bench, a man's deep in a book — no phone in his hands at all, which honestly you don't see every day. He's got that posture people settle into after an hour of reading, not ten minutes.

Out on the grass, two kids are playing a game they've clearly made up themselves — there's a ball involved but the rules look improvised, and the smaller one is losing and laughing anyway.

And along the path, an older couple with a little white dog — the dog's doing that elderly-dog trot where the walk is the whole event. Nobody's in a hurry anywhere in this frame. It's the park at the hour it works best."

Sample pair 3: opinion (Task 7 style)

Prompt: should cities make public transit free? Give your opinion.

Band 7:

"I think this is an interesting question. Free public transit has advantages and disadvantages. On one hand, free transit is good because more people will use buses and trains, and this is better for the environment. Fewer cars means less pollution. On the other hand, free transit costs the government a lot of money, and maybe the service will not be good quality. So I think cities should think carefully about this issue. Maybe they can make transit cheaper but not completely free."

Band 9:

"Yes — make it free, and I'd go further: cities should stop thinking of transit as something that needs to pay for itself. A bus fare isn't a business transaction, it's a tax you collect at the door — and an inefficient one, because the fare box costs money to run.

Here's the practical case. When my city ran a dollar-fare pilot on two routes, ridership jumped — and a chunk of those new riders visibly left cars at home. Now multiply that: school-run traffic drops, shift workers take jobs they couldn't reach, a family with two cars discovers one is enough. You're not losing fare revenue; you're trading it for road costs and parking lots.

The real objection is 'who pays' — and fair, someone does. But we don't ask who pays for the roads those drivers were on. We decided a century ago that streets are free at the point of use, funded by everyone. Transit is the same public good, one decision behind."

The verdict: the 7's "maybe cheaper but not free" is a position arrived at by refusing to take one. The 9 commits and builds a case you could disagree with — which is what scores.

What actually separates them

Read the three pairs again and the pattern is blunt: the 9s are not using better words. Both columns have accurate grammar. Four things separate every 7 from every 9:

Interpretation. The 7 reports; the 9 explains what the facts mean. "He is reading" vs. "the posture people settle into after an hour." The scene, the advice, the opinion — every 9 attaches a because.

Specificity. People, numbers, images. "Dollar-fare pilot on two routes." The 7s run on "some," "many," "very" — the strongest band-7 smell there is.

Rhythm. Every 7 sample has sentences of nearly identical length, front-loaded with "There is" and "I think." Every 9 follows a long sentence with a short landing ("Not yet."). That variety is what listenability measures.

Register. The 7s talk about the friend, about the question. The 9s talk to the friend, at the question. Advice sounds like advice; opinions sound held.

None of that is vocabulary — which is why word lists fail almost everyone who tries them. You can't memorize your way into a habit of attaching reasons.

Your this-week checklist

• Record one answer per day, any task type, out loud, timed
• Score each against the four criteria using the tables above
• Circle every vague quantifier — "some," "many," "very" — and replace it with a number or an image
• Add one "because" mechanism per answer; if a point has no reason, cut it
• Count your sentence lengths: if they're all 12–16 words, break the pattern
• By day 7: compare your first and last recordings — you'll hear the difference before you can explain it

Want these verdicts on your own recordings instead of ours? That's exactly what our evaluation does — band-level feedback on all four criteria → celpipuni.com/evaluation.

The deeper dive: why "more vocabulary" is the wrong lesson

Here's the opinion this post exists to argue: the most common study plan for CELPIP Speaking — memorizing advanced words — is aimed at the wrong target, and the samples above prove it. Hunt for impressive vocabulary in the 9 columns. You'll find "improvised," "at the point of use." That's about it — everything else is the ordinary English of a fluent person talking.

What the band-9 samples have cannot be memorized, only practiced: a reason attached to every fact, a specific where a generic would do, a rhythm that varies because the thinking varies. Uniform sentences aren't a style choice — they're what it sounds like when each sentence comes out of the same template. Varied rhythm happens when you're deciding what to say as you say it, which is the skill CELPIP Speaking measures.

So when people ask us "which vocabulary list should I study for a 9?" — the honest answer is: none, at least not first. Record an answer, transcribe it, fix the four things above. Two weeks of that beats two months of word lists.

Tools and resources

• The official CELPIP site — free sample prompts and the current task format. (Formats get revised; confirm there.)
• CELPIP Speaking: the complete guide — the four criteria plus the daily practice loop: celpipuni.com/celpip-speaking-guide
• Task-by-task samples — annotated full answers for the advice and personal-experience tasks: celpipuni.com/celpip-speaking-task-2-samples

Frequently asked questions

Are these real student recordings?

No — these samples were written for this post as teaching examples, and the verdicts are how our graders would score them. Real feedback needs your own recordings: that's what the evaluation is for.

How many words is a 60-second CELPIP answer?

Roughly 130–160 at a natural pace. Aim to fill the time with developed points, not padded ones.

Is band 7 bad?

No. A 7 maps to CLB 7 — enough for several Express Entry programs. It's a solid, functional level; the gap to 9 is just narrower than it looks.

Can I score 9 with a simple vocabulary?

Yes — look at the samples. Precision beats rarity in every criterion. "Boring is what survives" is five ordinary words and it's a band-9 sentence.

Does accent affect my band?

Intelligibility is scored; accent isn't a criterion. Clarity helps listenability, but no rubric line rewards any particular accent.

How do I know which band I'm at now?

Record an answer and compare it to the tables in this post — or get a baseline from a free mock test, then an evaluation once you're practicing seriously.

What to do next

1. Today: record one answer, any prompt, 60 seconds. Transcribe it word for word.
2. Score it against the four criteria using the verdict tables. Be brutal.
3. Fix the vaguest sentence, then re-record.
4. This week: get one evaluation to calibrate your self-scoring.

Related reading

• CELPIP Speaking: The Complete Guide to a 9+
• CELPIP Speaking Task 1: How to Give Advice That Scores 9
• CELPIP Speaking Task 3: Describe a Scene Like a 9
• CELPIP Speaking Task 7: The Opinion Playbook
• CELPIP Speaking Criteria: The 4 Things Raters Score

The bottom line

Band 7 speakers are accurate. Band 9 speakers are present — they attach meaning to facts, reach for the specific over the generic, and let their rhythm follow their thinking. None of that comes from a vocabulary list; all of it comes from recording yourself and fixing the same four things, daily. The samples are your rubric. Your recordings are the test.

Get your real band level with instant section scores — free mock test → Take your free mock test.