Word counter and readability checker.
Word count, reading time and a readability grade for any draft. The useful part is not the total — it is the average sentence length hiding underneath it.
By Mike Cecconello · Updated September 4, 2026
Word counter and readability checker
Runs in your browser. Nothing you paste is sent anywhere.
What you can do with it
- Check a draft against a client's length brief before you send it
- Find the sentences making a technical page unreadable
- Sanity-check reading time for a newsletter or landing page
- Compare a draft's readability against the pages already ranking
How it works, step by step
- 1
Paste the draft, not the outline
Readability is a property of finished sentences. An outline scores well because bullet fragments are short, which tells you nothing about the prose a reader will meet.
- 2
Read the average sentence length first
It is the number with the most leverage. Flesch reading ease is driven far more by sentence length than by vocabulary, so a page full of ordinary words can still score as graduate-level if the sentences run to forty words.
- 3
Target a band, not a number
Aim for plain English (roughly 60 to 70 reading ease, 8th to 9th grade) for most commercial content, and accept 50 to 60 for genuinely technical material. Chasing a specific score produces stilted writing.
- 4
Only then look at word count
Compare against the pages currently ranking for your query rather than a universal target. If the top results answer the question in 900 words, a 3,000-word version is padding, and padding is the thing readers and models both discard.
- 5
Fix the longest sentences, not the longest words
Split the worst three sentences and re-run. The score moves more than any amount of synonym substitution, and the writing gets better rather than simpler.
Why word count is the wrong target
Every content brief specifies a word count and almost none of them should. The number is a proxy that escaped its purpose: someone noticed that thorough pages tend to be longer, wrote “1,500 words” on a brief, and the writer padded to hit it.
What the brief meant was cover the question completely. Those are different instructions, and only one of them produces a page worth reading. A 700-word page that answers the query and stops is stronger than a 2,000-word page with the same answer and 1,300 words of preamble, because the padding is exactly what a reader skims past and what a model discards when it extracts a passage.
The one legitimate use of word count is comparative. If the pages ranking for your query all run to about 1,200 words, that tells you something real about what the query demands. It is a floor discovered from evidence, not a target invented in a brief.
The number that actually matters
Flesch reading ease is a two-variable formula: average sentence length and average syllables per word. Sentence length dominates. This means the most common readability problem is not vocabulary — it is writers joining three ideas with commas because each one felt too small to be its own sentence.
The fix is mechanical and it works. Find your three longest sentences and split each one. Re-run the count. The score moves further than any thesaurus exercise, and the prose improves rather than getting dumber, which is the failure mode of writing to a grade level.
Where this fits
Length and readability are the last checks, not the first. They tell you whether a draft is well written; they cannot tell you whether it was worth writing. That question is answered before the draft exists, by whether the query has demand you can realistically rank for — which is what a keyword plan is for, and why this tool sits in a set that starts with the plan rather than the prose.
Frequently asked questions
Is word count a ranking factor?
No. Google has said repeatedly that there is no minimum or ideal length. Longer pages correlate with better rankings mostly because thorough answers tend to be longer, not because length itself is rewarded. The causation runs through completeness.
What is a good readability score for SEO?
For most commercial content, a Flesch reading ease of 60 to 70 — plain English, around 8th to 9th grade — is a sensible target. Technical documentation legitimately sits lower. The score matters less than consistency: a page that swings between very simple and very dense reads as though two people wrote it.
How is reading time calculated?
This tool uses 225 words per minute, a common average for adult silent reading of general prose. It is an estimate: a dense technical page reads slower and a listicle reads faster, so treat it as an order of magnitude rather than a promise to your reader.
Does readability affect AI citations?
Indirectly, and through structure more than through score. Assistants extract passages they can lift cleanly, and a clear short sentence that answers a question is far easier to quote than the same fact buried in a subordinate clause. Optimising for extractability tends to improve readability as a side effect.
Why does the grade level disagree with other tools?
Because syllable counting without a pronunciation dictionary is a heuristic, and different tools guess differently. Treat the band as the signal and the exact number as noise. What is reliable is the direction it moves when you edit.
More free tools
- Free FAQ Schema Generator (JSON-LD, Article, Breadcrumb)
- Free Internal Link Checker: Audit Anchors on Any Page
- Free Keyword Clustering Tool: Group Keywords by Topic
- Free Keyword Gap Analysis Tool: Compare Two Keyword Lists
- Free robots.txt Generator With AI Crawler Rules
- Free Slug Generator: Clean URL Slugs From Any Title
- SEO Health Check: The Agency Audit Checklist
- SERP Analysis: How to Read a Search Result Page
Be the site the assistant names.
Add a domain and the first draft is written while you watch. Every publish is your decision, a click on the draft or a rule you set and can hold, and you can read the source or self-host it free.
Add a domain, it sets up your workspace