How to Score High on Describe Image — Without a Template
For years the advice was the same: memorise one Describe Image template, slot in a few numbers, repeat for every picture. The 2025 PTE refresh quietly killed that strategy. If your plan is still "recite the script," you're leaving points on the table — and possibly getting flagged.
Why templates stopped working
The post-2025 scoring is far better at spotting canned, repetitive language that doesn't actually engage with the content on screen. A response that opens with the same 40 memorised words every time signals one thing to the scorer: this person isn't really describing *this* image.
The engine rewards relevant content, natural fluency, and clear pronunciation. A rigid template hurts all three — it pads your 40 seconds with filler, it sounds robotic, and it crowds out the specific details that earn content marks. The students stuck at a plateau are almost always the ones still leaning on a script.
The mindset shift
Stop trying to say the *same thing* about every image. Start trying to say the *true thing* about this one, quickly and clearly. You don't need fancy language — you need accurate, specific observations delivered without long pauses.
A flexible 4-part method
Not a script — a thinking order. It works for charts, maps, processes, and photos alike.
- Open with what it is (1 sentence). "This bar chart compares smartphone sales across four regions." Name the type and the topic. One breath, then move on.
- Point to the headline (1–2 sentences). What jumps out? The biggest value, the clearest trend, the obvious contrast. "Asia leads by a wide margin, while Europe is the lowest."
- Add two real details. Pick two specific features and say them: a number, a peak, a category, a stage in a process. Specifics are what score — vague words like "various" and "different" do not.
- Close with a takeaway (1 sentence). A simple conclusion: "Overall, sales rise steadily from west to east." Done.
That's roughly 35–40 seconds of *relevant* speech with no memorised filler.
The 5 seconds before you speak
Use the short prep window deliberately:
- Find the single most obvious feature. That becomes your headline.
- Find two supporting details you can name confidently.
- Don't try to cover everything. Three solid points beats ten rushed ones.
What actually moves the score
- Keep talking. Fluency means no long pauses and no repeated restarts. A slightly imperfect sentence that flows beats a perfect one you stall on.
- Be specific. Numbers, labels, directions. Specifics earn content marks; generic phrases earn nothing.
- Pronounce clearly, at a natural pace. Racing to fit a memorised paragraph is why so many template-users score worse, not better.
- Don't fear silence at the end. If you've made four clear points in 30 seconds, stop. Trailing filler can pull your score down.
The one habit to build
Practise on images you've never seen, out loud, with a 40-second timer — no notes. The goal isn't a perfect answer; it's the reflex of seeing an image and immediately knowing your headline and two details. That reflex is what templates were faking, and it's what the new scoring actually rewards.
Kooroo's Describe Image practice scores you on the post-2025 criteria — content, fluency, pronunciation — and flags template-like, repetitive language so you can train the flexible version instead.
Frequently asked questions
Are Describe Image templates banned in PTE?
Not banned outright, but the post-2025 scoring penalises canned, repetitive language that doesn't engage with the image. Memorised templates tend to lower content and fluency scores rather than raise them.
How long should a Describe Image answer be?
Aim to speak for roughly 35–40 seconds of relevant content: a one-line opener, the main feature, two specific details, and a short takeaway. Quality and specificity matter more than filling every second.
What scores highest on Describe Image now?
Relevant, specific content (real numbers and labels), natural fluency with no long pauses, and clear pronunciation. Generic phrases and memorised intros work against all three.
