Free CELPIP Writing Mock Test Two timed tasks, both scored in minutes
Sit the writing component the way the test runs it: an email in 27 minutes, then a survey response in 26 minutes, on prompts you see only when the clock starts. When you finish, both answers are scored on the four official categories and you get an approximate overall writing level. Free, no account.
- Choose a set
- Task 1 · 27 min
- Task 2 · 26 min
- Two reports
Today's set
One Task 1 email prompt and one Task 2 survey prompt from 16 original prompts. Press start to see the first prompt and begin the 27-minute clock.
Free · No account · Uses your 2 free reports for today · Official timing and word counts
What the mock reproduces, and what it cannot
The official writing component is 53 minutes long: Task 1 · Writing an Email (27 minutes) followed by Task 2 · Responding to Survey Questions (26 minutes), both typed, both with a recommended length of 150–200 words shown next to a live word counter. This mock keeps those rules. The prompt is hidden until you press start, the clock runs without pause, the text box locks at zero, and Task 2 begins as soon as Task 1 ends. The prompts are original, written in the official format: a situation with three bullet points and a named reader for Task 1, a survey with two options for Task 2. The set changes every day, and you can draw another.
What it cannot reproduce is the rating. On the test, two trained raters rate each task on four categories and the two task scores are combined, 50 percent each, into the level on your report. Here, the AI scorer rates the same four categories, verifies every error it reports against your text, and applies rules that keep the estimate conservative; on 28 answers at known levels, 25 estimates were within one level. The overall figure shown after the mock is the lower rounding of the two task levels: a useful planning number, not a score.
How to use the 53 minutes
| Minutes | Task 1 (email, 27 min) | Task 2 (survey, 26 min) |
|---|---|---|
| 0–2 | Read the situation twice; note the reader and the tone; write one detail for each bullet point. | Choose the option you can support best; write two reasons and one example each. |
| 2–20 | Draft: greeting, purpose line, one paragraph per bullet point, closing line, sign-off. | Draft: choice and thesis, two reason paragraphs, concession and conclusion. |
| Last 5–7 | Check the bullet points (or the thesis and the concession), the tone, the word count, then the seven recurring errors: for/since, on/in/at, articles, fewer/less, comparatives, to-infinitives, comma splices. | |
Aim for 170 to 190 words in each task. It is long enough to develop every point and short enough to re-read once. Finishing Task 1 early does not add time to Task 2, so use the spare minutes to check rather than to stop.
What the two reports show
Each task gets the full report: the estimated CELPIP level with its CLB and CEFR equivalents, a rating for each of the four categories (Content / Coherence, Vocabulary, Readability, Task Fulfillment), a checklist of what the task asked for, every verified error with a correction and the rule, your strengths and three priorities, and a level-10 model answer on the same prompt. Read the two reports side by side: the category that is lowest in both is your priority, and the one that drops from Task 1 to Task 2 shows what fatigue costs you.
The approximate overall level is the floor of the mean of the two task levels. A 9 and an 8 give an 8; a 9 and a 9 give a 9. The official combination is not published, and raters see the whole test, so treat the overall figure as a direction, not a prediction.
Rules that make the mock worth taking
- One sitting, no pause, no dictionary, no spell-checker, no phone. The point is to see what you produce under the real constraints.
- Type in the box on this page rather than pasting from a document; the clock is the test.
- Take the mock once a week at most. Between mocks, rewrite the two answers with the corrections, score the rewrites one task at a time on the home page, and work through the tips for level 9 and above.
- Alternate formal and friendly email prompts over several weeks; tone is the Task Fulfillment factor most people lose without noticing.
After the mock
If both tasks are at your target level, your writing is ready and the remaining risk is time management, so keep taking a mock every week until the test date. If one task is a level lower, the Task 1 guide and the Task 2 guide cover the structure and the phrases for that task, and the sample answers show the same prompts at two levels. To see what the target level is worth in Express Entry, use the CELPIP to CLB calculator.
Frequently asked questions
Is this the real CELPIP writing test?
No. It is a practice test in the official format: the same two task types, the same time per task and the same recommended word counts, on original prompts written for this site. Official raters do not see it; the levels come from the AI scorer and are estimates.
How many mock tests can I take?
One a day per device. Each task uses one of the two free reports a day, and the limit resets at midnight UTC. Your drafts stay in your browser, so a mock interrupted by the limit can be scored the next day.
What happens when the clock reaches zero?
The text box locks, exactly as on the real test, and you move to the next step with what you have written. You can also finish a task early; the unused minutes are not carried over.
Can I take the mock on a prompt from a book or a course?
The mock draws its prompts from the site's bank so that the timing starts with the prompt hidden. To score an answer to a prompt from elsewhere, use the scorer on the home page and choose "I answered a different prompt".
Why is my Task 2 level lower than my Task 1 level?
It is common: a survey response needs two developed reasons with examples and a concession, and most test takers write it after 27 minutes of Task 1. The Task 2 report lists the exact passages to fix; the Task 2 guide shows the four-paragraph structure that keeps the level up.
Is there a spell checker during the mock?
No. The box here is a plain text area, and the scorer counts spelling among Readability errors, so you see what you really know. On the official test, type carefully and keep two minutes at the end to re-read.
Scorer: gpt-oss-120b (OpenAI), Cloudflare Workers AI; method and calibration on the scoring page. Timing and word counts: CELPIP Test Format & Scoring (53 minutes for the writing component) and the official free sample test (27 and 26 minutes, 150–200 words), read 20 October 2026. Independent study site; prompts are original and are not test questions.