Word Frequency Cloud

A word frequency cloud is a visual representation of text data where each word's font size corresponds to how frequently it appears in the source text. Words that occur more often are displayed larger and more prominently, making it easy to identify key themes, recurring topics, and dominant vocabulary at a glance. Unlike a basic word cloud image, our generator also provides exact word counts, percentage shares, a filterable frequency table, and multiple export formats.

The tool processes your text through a six-stage pipeline. First, it tokenizes the input into individual words using Unicode-aware pattern matching that handles multiple languages, contractions, and hyphenated terms. Next, it filters out common stop words (like "the", "and", "is") unless you choose to keep them. It then counts the frequency of each remaining word and sorts them by occurrence. Finally, a spiral placement algorithm positions each word on the canvas, scaling font sizes logarithmically based on frequency, with collision detection to prevent overlaps. Everything runs entirely in your browser.

No. The word frequency cloud generator runs 100% client-side in your browser. Your text is never uploaded, stored, or transmitted to any server. There is no sign-up, no account, and no data collection. This makes it safe to use with confidential documents, proprietary content, or sensitive survey responses.

Stop words are the most common function words in a language — articles, prepositions, pronouns, and auxiliary verbs like "the", "is", "at", "which", and "on". They appear frequently in almost any text but rarely carry meaningful content. Filtering them ensures your word cloud highlights the most significant and informative words rather than being dominated by filler. You can toggle stop word filtering on or off, view the full list of approximately 280 filtered words, and add your own custom exclusions for domain-specific terms.

Yes. The tool supports n-gram analysis in three modes: unigrams (individual words), bigrams (two-word phrases), and trigrams (three-word phrases). N-gram detection reveals meaningful collocations that single-word counts miss — phrases like "machine learning", "customer service", or "data science". For n-grams, the filter removes phrases that start or end with a stop word, preserving meaningful multi-word terms while discarding noise.

You can upload text files in .txt, .md (Markdown), .csv, .json, and .log formats. The tool reads the file content directly in your browser and loads it into the text editor for analysis. Files up to 5 MB are supported. You can also simply paste or type text directly into the input area.

The tool offers four export options. Download the frequency data as a CSV file (spreadsheet-ready with word, count, and percentage columns) or as a JSON file (structured data for programmatic use). Export the cloud itself as an SVG vector image that scales to any resolution. Or use the copy button to copy a word:count list to your clipboard for quick pasting into documents or chats. All exports are generated instantly in your browser.

Yes. The tool includes seven built-in color schemes — Ink, Sunset, Ocean, Forest, Monochrome, Rainbow, and Pastel — and three text orientation modes: Horizontal (all words left-to-right), Mixed (alternating horizontal and vertical for visual variety), and Vertical (all words rotated 90 degrees). You can also adjust the maximum number of words displayed and the maximum font size using sliders.

By default, the tool is case-insensitive, meaning "Apple" and "apple" are counted as the same word. This is useful for general text analysis where capitalization is incidental (e.g. sentence-start capitalization). Enable case-sensitive mode when capitalization is meaningful — for example, when "Apple" (the company) and "apple" (the fruit) appear in the same text and should be tracked separately. Stop word filtering always works case-insensitively regardless of this setting.

Font sizes are calculated using a logarithmic scale based on each word's frequency relative to the most and least frequent words. The most frequent word gets the maximum font size, and the least frequent gets the minimum. A logarithmic distribution (rather than linear) prevents smaller-frequency words from becoming unreadably tiny while still clearly showing the hierarchy. You can adjust the maximum font size with a slider; the minimum is fixed at a readable 13px.

Yes. In addition to the built-in stop word list, you can enter custom exclusion words or phrases in the "Custom Exclusions" field, separated by commas or new lines. Single-word exclusions remove any token matching that word. Phrase exclusions (multi-word entries) remove exact phrase matches. This is useful for filtering out character names in literary analysis, domain-specific filler in technical text, or any terms that are irrelevant to your analysis.

The cloud view renders an interactive SVG visualization where word size reflects frequency, with hover-to-highlight and tooltip features. The table view presents the same data in a searchable, scrollable list with rank number, word, exact count, a visual frequency bar, and percentage share. Both views use the same underlying frequency data — toggle between them to switch between visual exploration and precise data inspection. The SVG export is only available in cloud view; CSV, JSON, and copy are available in both.

Yes. The tokenizer uses Unicode property escapes that match letters and numbers across all scripts, including accented Latin characters, Cyrillic, CJK (Chinese, Japanese, Korean), Arabic, and more. However, the built-in stop word list is English-only. For non-English text, you can disable English stop word filtering and use the custom exclusions field to enter stop words in your target language. The frequency counting, cloud layout, and export features work identically regardless of language.