Keyword Density🔒 100% local
Word frequency analysis. Privacy-first, no server uploads, no tracking.
Related Tools
Reads your text twice over: once for a ranked keyword list with a fixed list of 80 common words removed, and once for two-word phrases that occur more than once. Density is worked out against the whole word count rather than a filtered one, and the report closes with three lines on where a keyword belongs on a page. No settings to pick; the text stays in the browser.
How to use it
- Paste the page copy you want to check into the box.
- Press Process. There are no options, the filter and the 15-entry list length are fixed.
- Read the keywords, then the repeated phrases, then the guidance lines at the end.
Worked example
JWT tokens carry signed claims. A JWT decoder reads the claims from a JWT without verifying the signature. Every JWT tool here runs offline, so a JWT never leaves the browser.becomes=== Keyword Density ===
Total words: 31
Unique keywords (non-stopword, 3+ chars): 18
Top keywords:
1. jwt 5× 16.13%
2. claims 2× 6.45%
3. tokens 1× 3.23%
4. carry 1× 3.23%
5. signed 1× 3.23%
6. decoder 1× 3.23%
7. reads 1× 3.23%
8. without 1× 3.23%
9. verifying 1× 3.23%
10. signature 1× 3.23%
11. every 1× 3.23%
12. tool 1× 3.23%
13. here 1× 3.23%
14. runs 1× 3.23%
15. offline 1× 3.23%
Repeated 2-word phrases: (none repeated)
SEO guidance:
• Aim for 0.5-2.5% density on primary keywords - higher reads as stuffing.
• Google's NLP now weighs semantic relatedness, not raw counts.
• Use your top keyword in title, H1, first 100 words, and one subheading.The same three-sentence sample, 31 words in total, so jwt lands at 5 of 31, or 16.13%, against 18.52% in the filtered count used by the plainer keyword density tool. The repeated-phrase section stays empty because no two-word pair shows up twice in a sample this short, and only 15 of the 18 distinct keywords are listed.
When to use it
- Deciding which phrase deserves the H1 and the title tag on a page that is being rewritten.
- Spotting a paragraph where one term is hammered hard enough that the copy reads like it was written around the keyword.
- Finding accidental repeated pairs, such as consecutive sentences that open with the same two words.
What it does not do
- The phrase section only lists bigrams appearing at least twice, so it is empty on anything short - a 300-word page will rarely produce a line there.
- Common words are stripped before keywords are counted but not before the total is counted, which pushes every percentage below what the other keyword density tool reports for the same text.
- The stopword list is baked into the tool, so a domain word that happens to be on it disappears from the ranking with no way to keep it, and no stemming is applied either.
- It measures repetition in the pasted text. It cannot tell you how a search engine treats that text, and the density figures it prints are not a ranking signal on their own.