Keyword Density Checker

How do you check keyword frequency and build a bulk keyword report?

Paste the text you want to inspect into Text to analyze. Enter your target list in Target keywords (one per line or comma-separated) to get a count and density for each term. The list accepts up to 50 distinct entries, including phrases. This answers a bulk keyword check for one document; it does not crawl a set of URLs or produce search ranking reports.

For a reproducible example, paste seo tools help writers seo tools improve drafts. There are eight counted words. Add seo, seo tools, and drafts as targets. Density uses occurrences ÷ counted words × 100, even for a multi-word phrase. It does not multiply the phrase count by the number of words in the phrase.

Frequency is the raw Count value; density makes that count relative to document length. Click Count in the target table to cycle through descending order, ascending order, and original list order. Targets that differ only in capitalization are deduplicated. A target missing from the text still gets a row with zero occurrences, which helps review a supplied list without relying on the top-keyword table.

Which words count toward the density percentage?

The standard tokenizer converts text to lowercase, keeps Unicode letters and numbers, and replaces most punctuation with spaces. It preserves internal apostrophes and hyphens, removes them from token edges, and discards tokens of one character. For example, I use on-device seo/tools produces four counted tokens: use, on-device, seo, and tools in that order. Add another counted word before expecting the main results to appear.

The main results require at least five counted words. If a document contains many single-letter words, its total can differ from a simple whitespace count. Use Word Counter for a separate text-length review, but keep this checker's displayed total as the denominator when verifying its percentages. Neither changing the minimum word length nor filtering stop words recalculates that denominator.

Filter stop words is checked initially and uses an English list. The initial minimum word length is two characters and the minimum frequency is one occurrence. The keyword table shows up to 20 entries in descending count order. Custom exclusions accept comma-separated or newline-separated words; they apply even when the built-in stop-word filter is disabled. These controls decide which terms appear, rather than removing those terms from the source text.

How do standard phrases differ from custom 20-word phrases?

The 2-word Phrases, 3-word Phrases, and 4-word Phrases tabs use the standard counted tokens. Each shows up to ten phrases appearing at least twice, or at your higher minimum frequency. A phrase is excluded when every token is a stop word. Stop words inside a mixed phrase remain in place; filtering does not join the words on either side of them.

The custom tab starts at five words and accepts lengths from 1 to 20. It uses a separate scan that retains stop words and single-character tokens. It splits sentences on periods, exclamation marks, and question marks when followed by whitespace or the end of the text. Standard phrase tabs do not make that sentence split, so their results can differ from custom results at the same length.

For example, paste we make a useful guide. we make a useful guide. and keep the minimum frequency at one. The custom five-word scan finds we make a useful guide twice, including the single-letter word. The standard tokenizer drops that letter, so it cannot report the same five-word sequence. Custom results require at least two occurrences and show up to 20 phrases. Use this view to locate repeated passages, then inspect whether they belong in the document.

What do the density colors and status labels mean?

The main keyword badges are green at 2% or below, yellow above 2% through 3%, and red above 3%. The target table distinguishes zero matches, below 1%, 1–2% labeled Optimal, above 2% through 3% labeled High, and above 3% labeled Stuffing risk. These are fixed interface thresholds applied to the calculated percentage, not a ranking measurement or an evaluation of writing quality.

In the eight-word example, 25% triggers the highest target warning because the denominator is small. That warning alone does not explain whether the sentence is useful or repetitive. Read the surrounding text before editing. Click a keyword row to highlight it in the preview, or use multi-highlight to mark up to five leading keywords together. Pair the counts with Readability Checker when reviewing how the draft reads.

How do you compare two texts and export the results?

Enable Compare with a second text and paste the other document into Text B (for example a competitor page). The main input is Text A. Comparison supports single words, two-word phrases, and three-word phrases, with up to 20 results for each selected length. Results sort by the largest absolute density difference. Delta is density B minus density A: 5% in A and 8% in B gives +3 percentage points.

A term gets missing in A when its count in A is zero and its count in B is at least two. This is a literal occurrence check within the supplied content. The tool does not fetch a page from its URL, check keyword positions in search results, or identify topics that a ranking system expects. Paste comparable portions of both documents if you want the percentages to describe similar content.

Copy Keywords copies the current tab's rows as tab-separated text. Download CSV saves the standard keyword and phrase tables plus available target, custom phrase, and comparison sections. The CSV uses UTF-8 with a byte order mark. Analysis runs in browser code; Upload .txt reads a selected local text file with FileReader, and CSV download uses a locally created Blob.