AI Hashtag Generator — from your caption, free

This generator ranks hashtags from your own text— it never guesses “trending” tags or invents reach numbers, because no free, stable, citeable source for per-hashtag volume exists. Everything runs in your browser as a pure, deterministic calculation: the same caption and settings always produce the same tags, and no network request fires.

1. Extract candidate phrases.Your text is scored by two independent, peer-reviewed statistical extractors — YAKE (Campos et al. 2020), which weighs each word's position, casing, and neighbour diversity, and RAKE (Rose et al. 2010), which scores phrases by co-occurrence degree over frequency. This is the same engine that powers the AI Keyword Extractor on this site. Phrases both methods surface are marked both — the highest-confidence keywords.
2. Normalise to hashtag tokens. Each phrase is NFKD-normalised to strip diacritics (so café becomes cafe), stripped of every non-alphanumeric character, and collapsed into a single token (butter cake → ButterCake). Tokens shorter than two characters and pure-numeric tokens are dropped. CamelCase capitalises each word boundary for screen-reader legibility; lowercase flattens everything.
3. Augment with an optional niche pack. If you pick a niche, a small curated bank of established broad community tags is blended in so a thin caption still yields a usable set. These curated tags are always scored below your own extracted tags, so your content ranks first, and they are labelled Curated for transparency.
4. Classify and rank. Each tag is labelled Broad (a single common word), Niche (a multi-word compound), or Branded (a CamelCase or all-caps proper noun detected in your text). The list is sorted with your extracted tags first, then by relevance. Case- insensitive duplicates are merged, keeping the higher score.
5. Apply platform rules.The count is clamped to the platform's real cap, then the block is built and measured. Every block is independently re-counted and re-measured before it is shown, so the displayed count never exceeds the cap and nothing is silently truncated — anything trimmed appears in a “Removed” note.

The platform caps applied are Instagram 30, TikTok 30, X (Twitter) 3, Facebook 30, LinkedIn 5, YouTube 15. Each is sourced from the platform's own help centre, linked in Sources & references below.

• Instagram Help Center — Add hashtags in your posts (max 30; 2,200-char caption)
• TikTok Business Help — video captions (2,200-char caption)
• X (Twitter) Help — How to use hashtags (280-char post)
• LinkedIn Help — Hashtags on LinkedIn (3–5 recommended)
• YouTube Help — Add hashtags to video descriptions (max 15)
• Campos et al. (2020) — YAKE! keyword extraction (ranking method)
• Rose et al. (2010) — RAKE keyword extraction (ranking method)

Platform limits were last cross-checked against each official help centre on 2026-06-09. The curated niche packs are an editorial convenience bank of broad community tags, not a reach metric — your own extracted tags always rank first.