# Free LinkedIn Post Grader

Paste a draft and find out whether there is anything in it worth reading. Every finding quotes the line from your post that produced it, so you can argue with any of them.

The whole report appears on screen without an email address or an account.

## Knowing what is missing is the easy half.

A grader can tell you the post has nothing in it that could only be yours. It cannot supply the missing part, because it has never met you and knows nothing about your week. Oiti reads the posts you have already published and holds what you know about your work in a knowledge base, so the next draft turns up with your material already in it. Start the 7-day free trial at [ghostwriting-ai.com](https://www.ghostwriting-ai.com).

## How to grade a LinkedIn post

1. **Paste the draft.** Drop the post into the box before you publish it. Anything from 80 to 3,000 characters works, which covers everything LinkedIn will let you post.
2. **Read the grade and the five checks.** Every finding quotes the line from your post that produced it, so you can disagree with any of them on the spot.
3. **Fix the material, then grade it again.** The fixes are instructions about your own draft rather than sentences to paste, and they stick to things you can say in public. Change what you agree with and run it again.

## What you get

- One grade from A to D, with the verdict behind it
- Five checks: anchor, hook, specificity, focus, close
- A line quoted from your own post behind every finding
- Phrasing patterns that read as machine written, each with the line it came from
- Two or three fixes, written as instructions rather than as copy to paste
- A plain-text report and a downloadable score card
- Ten grades a day, free, with no signup and no email

## What this checks that other graders do not

Most graders read the shape of a post. Length, line breaks, whether the first line has a number in it, whether there is a question at the end. Those are real things, and every one of them can be measured without understanding a single sentence. That is a large part of why every tool measures them.

Shape is not what makes a post worth reading. You can hit every formatting rule on the list and still publish something no reader could tell apart from a thousand other posts that week.

So the first check here is a different kind. It asks whether the post contains anything only you could have written: one concrete detail, a moment, a named person, a real number attached to a real event, a scene you were in, or an opinion you can defend that plenty of people would argue with. Twelve years of experience is not it, and neither is "consistency compounds". The test is whether the sentence would still be true in somebody else's mouth.

When that is missing, nothing else rescues the post and no rewrite repairs it. The hole is in the material, so the only fix is to go and get some.

## The five checks

Each one comes back with a band, a line quoted from your post, and one sentence on why.

- **Anchor.** Is there one concrete, only-yours detail in here? The make-or-break check, and it counts double. Present, thin, or missing. Bands: Present, Thin, Missing.
- **Hook.** LinkedIn folds a post behind "see more" at roughly 140 characters or two lines on mobile, where most of your readers are. Does the opening earn the tap inside that space, or does it spend it on a preamble? Strong, workable, or weak. Bands: Strong, Workable, Weak.
- **Specificity.** Names, numbers, timeframes, and things that happened, weighed against abstractions that would be equally true for anyone. Concrete, mixed, or abstract. Bands: Concrete, Mixed, Abstract.
- **Focus.** One thesis, or a main point with three generic detours bolted on. Single idea, drifting, or scattered. Bands: Single idea, Drifting, Scattered.
- **Close.** Does the ending land on what is at stake in this post, or is it a recycled sign-off that would fit under anything? Earned, flat, or recycled. Bands: Earned, Flat, Recycled.

## How the grade is worked out

Each check earns two points at the top band, one in the middle, zero at the bottom. The anchor counts double, so twelve is the maximum.

| Points | Grade | Verdict |
| --- | --- | --- |
| 11 to 12 | A | Ready to post |
| 8 to 10 | B | Close |
| 5 to 7 | C | Needs work |
| 0 to 4 | D | Start over |

One override sits on top of that table. If the anchor is missing, the grade cannot climb above C however good the rest of the post is. A well-hooked, tightly focused, beautifully closed post about nothing is still a post about nothing, and not one of the graders we tested said so: they scored the draft and then offered advice on the hook.

The bands are picked by a language model reading your post. The arithmetic happens in our code, against the table above. That split is the whole design. Models are reasonable at judging whether a sentence is concrete and unreliable at producing a number that means anything, so we only ask for the part they can do. A score of 6.8 out of 10 has the texture of precision and nothing behind it.

## What this LinkedIn post analyzer will not do

- **No rewrite of your post.** No suggested hook, no replacement sentence, nothing you could paste straight into LinkedIn. That is a real limitation and it is on purpose. Running the same draft through the other graders, two of them handed back rewritten copy containing a statistic and a result the author had never claimed and we could not find anywhere. A tool that writes no prose has nowhere to put an invented number.
- **No engagement prediction.** Nobody can do this. The tools that stamp a viral score on your draft are guessing with a straight face.
- **No AI detection.** The report flags phrasing patterns that read as machine written and quotes the line each one came from, because those patterns are worth catching whoever typed them. That is a craft observation rather than a detection score, and there is no percentage, no named detector and no bypass claim anywhere on this page.
- **No knowledge of your writing.** It knows what a post with substance looks like. Recognising what one of yours sounds like takes your published posts, and that is the paid product.

## Why a good-looking post still flops

The usual failure is not a badly written post. It is a well-written post with nothing inside it.

Picture the draft. Short paragraphs, plenty of white space, a question in the first line, three tidy points in the middle, then a closing line that sounds like a conclusion followed by a question aimed at the comment box. Every formatting rule met, top to bottom.

Nothing was in it, so the formatting had nothing to carry. A reader gets two lines in, recognises the shape, and scrolls, because they have read that post before under three other names.

A shape-based grader handles this case worst, because on the surface the post is doing everything right. It is also the most common way a draft fails, which is an awkward combination.

## Getting the material, not just the grade

A grader can tell you the anchor is missing. It cannot hand you one, having never met you and knowing nothing about your week.

Oiti works from your material instead. It reads the posts you have already published and picks up how you open, what you are blunt about, the phrases you keep reaching for. It also holds what you know about your business in a knowledge base, so drafts turn up with your specifics in them rather than a placeholder where the specifics should go.

Seven days free, and nothing on this page sits behind it.

## Frequently asked questions

### Is it actually free?
Yes. No signup, no email address, no account, and the whole report appears on screen. There is a cap of ten posts per person per day, and it exists to keep the model bill survivable rather than to push you toward an upgrade. Worth saying because it is not the norm here: the tool currently ranking first for this search asks for your email after you click the button, on a page whose description promises no signup.

### Why a letter grade instead of a score out of 100?
Because the number would be invented. A language model can tell you fairly reliably whether a sentence is concrete or abstract. Ask the same model for 7.4 out of 10 and you get a figure with the texture of precision and nothing underneath it. So the model only picks bands, and our own code turns the bands into points and a letter using the table on this page. If you can check it, you can trust it.

### What is an anchor?
One concrete detail in the post that could only have come from you. A specific moment, a named person, a real number attached to a real event, a scene you were in, or an opinion you can defend that plenty of people would argue with. General advice is not an anchor however true it is, and neither is background about yourself. The test: would the sentence still be true in somebody else's mouth? If yes, it is padding.

### Can it roast my post?
Not in the comedian sense. The grade gets blunt when the draft deserves it, but there is no roast mode and no persona. A joke about your writing is entertaining and tells you nothing about what to change, and the tools doing that well are doing a different job. If what you wanted was the harsh read rather than the funny one, this gives you that.

### Will it rewrite the post for me?
No, deliberately. The fixes are instructions about your own draft, along the lines of "put the number you are allowed to share into the second line" or "cut the last sentence, it repeats the post". One exception: if a sentence further down your post would work better as the opening, the report says so and quotes it back, because moving your own line is not the same as writing you a new one. Two of the other graders we tested handed back rewritten copy containing facts the author never claimed, which is the last thing you want from a tool you are about to trust with something going out under your name.

### Does it check whether my post sounds AI generated?
It flags the phrasing patterns that read as machine written and shows you the exact line each one came from, so you can judge the call yourself. It is not a detector, it produces no detection score, and it makes no claim about any specific tool. Those patterns are worth cutting whether a model wrote the post or you did.

### What happens to the post I paste?
It goes to our AI provider to be read and the report comes back to your browser. On our side nothing is saved: no account, no database row, no public gallery, and error logs that record the request size and status rather than what you wrote. What the AI provider retains is governed by their terms. The safe habit with any tool like this one, ours included, is not to paste anything confidential.

### How long should a LinkedIn post be?
Long enough to carry one idea and the specific detail that makes it yours, and no longer. You will find confident ranges quoted around the web, and they disagree with each other because each one comes from a different sample of different people's audiences. This tool does not score length at all. It scores whether the post has anything in it.

## More free LinkedIn tools

- [LinkedIn post readability checker](https://www.ghostwriting-ai.com/tools/linkedin-post-readability-checker) Six published readability scores, each one showing the arithmetic behind it.
- [LinkedIn post humanizer](https://www.ghostwriting-ai.com/tools/linkedin-post-humanizer) Strip the machine-written phrasing out of a draft you already have.
- [LinkedIn post generator](https://www.ghostwriting-ai.com/tools/linkedin-post-generator) Write a first draft from a topic, a format and a tone.
- [LinkedIn Wrapped](https://www.ghostwriting-ai.com/tools/linkedin-wrapped) See how your last 12 months of posts did, from LinkedIn's own analytics export.
- [LinkedIn text formatter](https://www.ghostwriting-ai.com/tools/linkedin-text-post-formatter) Add bold, italics and clean line breaks that survive the paste.

## Why we built it

The post grader is a free entry point to Oiti, the AI content engine for LinkedIn. This tool reads one draft and tells you what is missing from it. Oiti works from everything you have already published, the way you actually talk, and the material you feed its Knowledge Base, so the draft turns up with your specifics already in it. Seven days free. See [ghostwriting-ai.com](https://www.ghostwriting-ai.com).