ChatGPT can comment on your Writing Task 1 email, but a chat reply is not a CELPIP score. CELPIP says that each Writing performance is rated by four to six trained raters who work independently of one another (CELPIP, test results and scoring, as of November 4, 2026). A chat assistant is useful for grammar, spelling, wording and clarity, and you can check each of those comments against your own text. Any score it gives is the tool’s opinion, so test the feedback before you act on it.
How does CELPIP score Writing Task 1?
CELPIP rates Writing with trained human raters, in four dimensions. According to CELPIP, raters must certify after a training program, an online system assigns tests to raters at random, and the test taker stays anonymous. CELPIP defines a performance as your responses to all the tasks in a component, so the raters assess your Writing as a whole, and a CELPIP Writing level covers both tasks.
Each rater assigns a level in each dimension by finding tangible evidence in your writing that matches written descriptors, and every dimension has five performance levels. If the raters disagree, a reference rater is assigned automatically. The component score is derived from the ratings by dimension and then transformed into a CELPIP level by rules that English language experts established. CELPIP lists these factors under each Writing dimension, and its Writing study pack pairs each dimension with one guiding question:
| Dimension | Factors CELPIP lists | Question in CELPIP’s Writing study pack |
|---|---|---|
| Content/Coherence | Number, quality and organization of ideas; examples and supporting details | How well are your ideas organized and developed? |
| Vocabulary | Word choice; suitable use of words and phrases; range of words and phrases; precision and accuracy | What is the range of your vocabulary and can you use it naturally? |
| Readability | Format and paragraphing; connectors and transitions; grammar and sentence structure; spelling and punctuation | How easy is it to read and understand your response? |
| Task Fulfillment | Relevance; completeness; tone; word count | How well did you follow the instructions? Did you use an appropriate tone? |
Source: CELPIP, test results and scoring and CELPIP’s Writing study pack, read November 4, 2026.
Is the CELPIP pass or fail? explains how CELPIP reports levels.
Why does the difference between feedback and a score matter?
A chat reply is text generated in answer to what you pasted. A CELPIP level is the end of a different process: independent ratings by dimension, then transformation rules set by language experts. On the screen the two can look alike, with a number and a few comments, but they do not carry the same information.
A number from a chat assistant looks precise, yet you cannot trace it to the written descriptors for the five performance levels in each dimension. It is the tool’s reading of your text. If you book a test date or decide to retake because of that number, you are deciding on one opinion. Used as feedback, the same reply is useful: you read each comment, find the line it points to, and decide whether it is right.
Who can give you feedback on a Writing draft?
You have three sources of feedback, and each one answers a different question. You can read your own draft against CELPIP’s four guiding questions, after a break so that the text looks new to you. A chat assistant can check grammar, spelling, wording and clarity quickly, and every suggestion can be verified in your text. A teacher can read the draft as a whole and explain what to change first.
Express Prep (expressprep.ca), which offers preparation for the CELPIP test, says in its terms that members of a program with live sessions may submit their Writing and Speaking for feedback, and that the feedback is subject to the availability of its mentors (terms read November 4, 2026). Self-study or a coach lays out how to decide between these options.
What does a chat assistant do with a Task 1 email?
CELPIP’s General Overview study pack describes Task 1 as an email about an everyday matter, written in 27 minutes and 150 to 200 words, in which you are usually asked to do around three things, in a tone that is neither too formal nor too informal (as of November 4, 2026). Here is a practice example with a trimmed draft.
An original practice prompt: Write an email to your building manager. Say that the heating in your unit stopped working, explain what you have already tried, and ask when it can be repaired.
Hey Dube,
The heating in my apartment stopped working on Monday and the rooms are very cold. I turned the thermostat up and checked the breaker, but nothing changed. Please tell me when someone can come to fix it, I have a baby at home.
Thanks for your help, Sam
The draft is trimmed to show feedback; a full response is 150 to 200 words. Paste it into a chat assistant with the prompt and ask for comments. Whatever the tool says, you can check three things yourself. The comma in “Please tell me when someone can come to fix it, I have a baby at home” joins two sentences, and a full stop or a conjunction fixes it. “Hey Dube” is casual for a message to a building manager, and the pack asks for a tone that is neither too formal nor too informal. The draft covers all three things the prompt asks for. Length is a separate check: this trimmed version is far below the 150 to 200 words a full response needs.
What a chat reply cannot know is just as concrete. It does not know how long you took unless you tell it. It cannot say which of the five performance levels each CELPIP rater would match to this draft in each dimension. And it cannot know whether you could write this text alone on test day, where the Writing screen offers basic editing tools, an automatic word count and a spell checker. For a paragraph map of the email, see how to structure Writing Task 1 and Task 2.
How can you test any AI feedback before you trust it?
Run these five checks on any tool, with your own draft and the prompt:
- Paste the same draft twice, in separate chats, and compare the comments and any score.
- Ask the tool to quote the line of your draft behind each comment, then find that line yourself.
- Request comments under the four CELPIP dimensions and check that each one points to something in your text, not to general advice.
- Ask for a rewrite and read it next to your draft. Check that it keeps your tone and vocabulary and stays between 150 and 200 words. CELPIP’s Writing study pack notes that even a revised sample response cannot be rated, because it was created for study purposes only.
- Treat any score as an opinion, and compare it with the reading of a second person.
Writing templates covers how a fixed outline fits with all of this. The same study pack reminds readers that studying a package and learning English are two different things. One practical job for an AI tool is to find the mistake you keep repeating, so that you can fix it in your own writing.
FAQ
Can I ask ChatGPT to use the CELPIP Writing dimensions?
Yes. You can paste the four dimensions and their factors from celpip.ca and ask for comments under each heading. The reply is then a structured opinion that uses CELPIP’s labels, not a rating, because CELPIP raters match your text to written descriptors for five performance levels in each dimension and work independently of one another.
Should I copy the version that ChatGPT rewrites?
Use it as a study text. A rewrite shows other ways to say something, but on test day you write alone, in 27 minutes, with a word count on screen. Pick two or three phrases you would really use, then rewrite your own draft without looking at the other version and compare the two.
Can ChatGPT check my word count?
Count the words yourself in a word processor. The task asks for 150 to 200 words according to CELPIP’s overview pack, the test screen has an automatic word count, and CELPIP’s tips page says your score may suffer if you are significantly over or under the limits. A tool’s count is one more figure to verify.
How long does feedback from a teacher take compared with a chat reply?
A chat reply arrives while you wait. Feedback from a person takes longer, and how much longer depends on who gives it. In the terms of Express Prep, the usual response time for Writing and Speaking feedback is 48 hours during normal business hours, and the terms reserve the right to answer within 48 to 72 hours (terms read November 4, 2026). Plan the review cycle with a gap like that in mind.
What should I do if two tools give different scores?
Set both numbers aside and go back to the four questions that CELPIP’s study pack asks about Writing. Mark the dimension where the two tools disagree most, reread your draft for that dimension only, and ask a person to read it. The disagreement tells you where to look, not which number is right.
The bottom line
- A chat reply is feedback on your draft, not a CELPIP score: CELPIP says each Writing performance is rated by four to six trained raters who work independently (celpip.ca, as of November 4, 2026).
- CELPIP lists four Writing dimensions: Content/Coherence, Vocabulary, Readability and Task Fulfillment.
- A chat assistant can help with grammar, spelling, wording and clarity, and each suggestion can be checked against your own text.
- Test any tool by running the same draft twice, asking for the line behind each comment, and comparing a rewrite with your draft.
- Teacher feedback takes longer than a chat reply, so plan the review before your test date.
Express Prep offers preparation for the CELPIP test, including a 7-day trial of its simulator. The trial becomes a paid monthly subscription unless you cancel before day 7; current price and terms: Express Prep trial page.
Published by Express Prep, which offers preparation for the CELPIP test. Last updated November 4, 2026.