JURASSIC ENGLISH · AI SPEAKING PRACTICE
A typed IELTS-style Speaking practice mock (Parts 1, 2 and 3), assessed by AI against the text-supported public Speaking criteria. See your own words as evidence, an honest uncertainty range, and the one thing to fix next. Pronunciation is not assessed from typing.
Built for university applicants, graduates and working professionals preparing for IELTS. Practice credits from 79,000₫ (1 credit), or 54,900₫ per credit in the 10-credit pack. Practice feedback only, not an official IELTS score. We do not store your answers on our servers; how the AI provider handles them is explained before you start.
Type your answers to Parts 1, 2 and 3. Work through the full practice mock at your own pace.
Each criterion is scored with a quote from your answer that shows why, and a confidence level.
Your report names your clearest next priority with your own words as evidence, what you already do well, and one practice action.
Illustrative example of a typed sitting from the demo scorer, not a real student's result.
Compare all three reports: Basic, Intermediate, Advanced →
This report is based on typed answers. We can analyse vocabulary, grammar, response structure and idea development, but pronunciation and full spoken fluency require a recording.
Fluency & Coherence and Grammatical Range & Accuracy are within half a band of each other in this sample; Fluency & Coherence is the most likely next priority (lowest band first, then the stronger evidence).
Take your sentence “What I admire most is her calm, because when I failed my first English test she did not criticise me.” and say it again as three linked sentences: your point, a reason (because…) and an example (for example…).
| Criterion | Band | Evidence (your words) |
|---|---|---|
| Fluency & Coherence | 6.5 | “What I admire most is her calm, because when I failed my first English test she did not criticise me.” |
| Lexical Resource | 8 | “That changed the way I think about failure, so I would say she is the most influential person in my life.” |
| Grammatical Range & Accuracy | 6.5 | “She grew up in a small village near Nam Dinh and never finished school, however she taught herself to read by borrowing newspapers from neighbours.” |
| Pronunciation | – | (not assessed: typed answers; pronunciation needs a recording) |
A band is only shown with a quote from your own answer that supports it.
You see a best-fit band with its likely range. When the evidence is thin, the range gets wider and we say why.
A second review pass may hold or lower a band, never raise it, so scores are not inflated.
Answers under 20 words are not scored. A guess dressed up as a band helps nobody.
Younger learners build IELTS Speaking skills with our teachers in the Jurassic English Young Learners programme: 1-to-1 and small-group coaching, with regular progress reports for parents.
Run IELTS-style Speaking mocks under your own brand and give every student an evidence-based report. We are opening a small pilot with licensed Vietnamese centers: your center stays in control of student data, and the pilot starts with adult learners.
Buy credits, get one access code, and take a full practice mock whenever you are ready. A Basic report uses 1 credit, Intermediate 2 and Advanced 3; the 3- and 10-credit packs cover any of them. Credits are valid for 12 months and are returned automatically if our scoring fails. There are 5 practice question sets; each new mock uses a set you have not taken yet, where possible.
It is AI practice feedback, not a test result. Only an IELTS test can award an IELTS score. This tool is not endorsed by IELTS, IDP, British Council or Cambridge, and it must not be used for grading, placement, admission or ranking.
Band targets are training goals, not guarantees. Every estimate shows its evidence and uncertainty.
Yes, and every report says so.
When there is no audio, or several criteria have low confidence, we widen the range rather than pretend to be precise.
No. Typed answers give a text-based practice estimate from vocabulary, grammar, response structure and idea development. Pronunciation and full spoken fluency need a recording, so a typed report shows pronunciation as not assessed and says so on the report.
The public practice flow currently uses typed answers. Spoken answers are limited beta for invited users. For those invited, If you type your answers, nothing is recorded. If you answer by speaking, each spoken answer goes through a processing service we run on Vercel in Singapore to Microsoft Azure Speech (Singapore), only to write down what you said and to measure pronunciation and fluency. We do not store the recordings, the Vercel service keeps no audio, and Microsoft processes them under its own terms. You can withdraw consent or ask us about your data at any time.
For adult learners: 1 credit for 79.000₫ (a Basic report), 3 credits for 199.000₫ (one Advanced report, or Intermediate + Basic, or three Basic reports), or 10 credits for 549.000₫ (up to 10 Basic reports, or a mix of report types). A mock uses its credits when you start it. Credits last 12 months, unused credits can be refunded within 14 days, and they are returned automatically if our scoring fails. Pilot terms for centers are agreed with each center.