Will your first 3 seconds hold viewers — or get scrolled past?

Paste the opening lines of your TikTok, Reel or Short and get a free 0–100 hook score, hook-type detection, per-line verdicts and rewrites — before you spend a post finding out the hard way.

  • Your script never leaves the device
  • No sign-up · no install · no account
  • Free & unlimited — no daily credits
  • Fixed public rubric, not a black box

Analyse your hook

0 / 2000 characters One line per opening line. Up to 12 lines, 2,000 characters. The first line counts most — it is the line viewers judge.
The rubric is the same on every feed — the first 3 seconds decide retention everywhere.
Free, unlimited, private. No daily credit limit, no account, no uploads — analyse as many hooks as you like.
Try a sample hook: See what a strong, a weak and a workable opening score — and why.

Paste your opening lines above — or load a sample to see how the scoring works.

    Take the report with you. Copy it into your notes or script doc, or download it as a text file.

    Privacy you can verify, not just trust

    Most hook analyzers are apps or AI services: paste a video link, wait for a server, and spend one of your 5 free credits a day — and your script sits on their server. This one can't: everything you paste is analysed in your browser, on this device. Your script never leaves the device — the only third-party requests are the site's own analytics, which count page views — and it uploads nothing, ever. Don't take our word for it:

    New cross-origin requests during use: 0
    1. Press F12 to open developer tools, then the Network tab.
    2. Paste a hook and analyse it. Watch: apart from the site's analytics, nothing new is requested — no uploads, and no script text ever leaves your device.
    3. Extra credit: disconnect your internet after this page loads — pasting, analysing and copying the report still work.

    How to test a hook before you post it

    The first three seconds decide most of your retention, and retention decides distribution. TikTok's own data says 63% of its highest click-through videos hook viewers within the first three seconds, and Instagram's head has said viewers decide on a Reel in about 1.7 seconds. The fix is to test the opening lines before you film — not after the post underperforms. When you paste a hook here, your browser:

    1. Names the hook type — question, list, pattern interrupt, bold claim, story or promise — or calls it a plain statement when nothing in it makes a scroller stop.
    2. Scores it on five factors — hook type, curiosity gap, specificity, clarity and 3-second fit, 20 points each — with the first line weighted at 60% of the overall score.
    3. Writes the fixes in plain English — the top 3 fixes across all lines, a verdict per line, and up to 3 suggested rewrites per line, built from your own words: cut the greeting, ask it as a question, trim to the budget, add a number.

    What makes this analyzer different

    The pages that top the results for this search are mostly paid apps, credit-capped AI services, or comparison articles. Hooky is an iPhone app with a $14.99-a-month subscription; ShortsIntel gives you 5 free lookups a day and sends your video to its servers; ViralFaceless caps you at 10 free grades a day. This page is the only free, fully client-side hook analyzer we found that runs in the browser with no install, no account, no daily credit limit and no uploads, scores the script against a fixed public rubric you can read on this page, and shows a live network-request counter so you can verify the privacy claim yourself.

    When is this the right tool — and when is it the wrong one?

    This is the right tool the moment the script exists and the camera has not rolled yet: paste the opening lines, fix the weak ones, and film the winner. It is the wrong tool for judging a finished video — it cannot see your visuals, hear your delivery, or know your audience — and it is the wrong tool for predicting views. A strong hook on a weak idea still underperforms; use the score to fix the words, then judge the video by how real viewers respond.

    Honest limitations

    It reads words, not video

    Your visuals, delivery, editing, music and audience are invisible to this tool. A great line delivered flat will still lose viewers — and a mediocre line delivered brilliantly can still win.

    The score is a guide, not a prediction

    No tool can guarantee views. The score says whether your opening lines follow the patterns that hold attention — the idea itself, the niche and the timing still decide the outcome.

    Speaking pace varies

    The 3-second budget assumes about 2.5 words per second. Fast talkers fit more in; slow, deliberate delivery fits less. Treat 7–9 words as a target, not a hard limit.

    Rewrites are starting points

    Suggested rewrites are template-based, built from your own line — not AI output and not finished copy. Edit them in your own voice before you film.

    English only

    The rubric's word lists (weak openers, filler, curiosity words) are English. Hooks in other languages will score lower than they deserve — the type detection and word-count checks still work, but the vocabulary checks do not.

    It scores the opening only

    Retention after the hook — mid-roll re-hooks, pacing, payoff delivery — is a different problem. This tool fixes the first 3 seconds; the rest of the video still has to earn the watch.

    Frequently asked questions

    Does my script upload anywhere?

    No. Everything you paste is analysed in your browser and never leaves the device, and the only third-party requests are the site's own analytics, which count page views. You can verify it yourself with the network-request counter in the privacy section below the analyzer.

    What does the 0–100 score actually measure?

    Five factors, 20 points each, against a fixed public rubric: hook type (question, list, pattern interrupt, bold claim, story or promise beats a flat statement), curiosity gap (questions, secrets, mistakes, open loops), specificity (numbers, timeframes, amounts), clarity (no weak openers, filler or vague words) and 3-second fit (7–9 words is the spoken budget). The first line counts most, because it is the line viewers actually judge.

    Why is the first 3 seconds the budget?

    Because that is when viewers decide. TikTok's own data says 63% of its highest click-through videos hook viewers within the first three seconds, and Instagram's head has said viewers decide on a Reel in about 1.7 seconds. At a natural speaking pace of roughly 2.5 words per second, about 7–9 words fit in those 3 seconds — so the payoff has to land inside the first nine words, not after them.

    Will a high score guarantee views?

    No, and no honest tool can promise that. The score is a guide to whether your opening lines follow the patterns that hold attention. It cannot see your visuals, delivery, editing, music or audience, and it has no view or retention data. A strong hook on a weak idea still underperforms — use the score to fix the words, then judge the video by how real viewers respond.

    What hook types does it detect?

    Six: question hooks, list hooks (a numbered promise such as '3 ways…'), pattern interrupts ('stop scrolling', 'wait', 'don't'), bold claims ('the best', 'nobody tells you'), story hooks ('I tried…') and promise hooks ('how to…'). A flat statement with none of those patterns scores lowest on hook type, because nothing in it makes a scroller stop.

    Does it work for TikTok, Reels and Shorts?

    Yes. The rubric is the same for all three — the first 3 seconds decide retention on TikTok, Instagram Reels and YouTube Shorts alike. Pick the platform above the analyzer and the report is labelled for it; the scoring does not change, because the attention problem is the same on every feed.

    Are the rewrites written by AI?

    No. The rewrites are template-based suggestions built from your own line: cutting a weak greeting, asking the line as a question, trimming it to the 3-second budget, or adding a specific number. They are starting points to edit in your own voice, not finished lines — and nothing you paste is ever sent to a server to generate them.

    What does this tool NOT do?

    It does not analyse video files or links — paste the script text instead. It does not predict views, watch time or virality, it does not see your visuals or hear your delivery, and it does not write your whole script for you. It scores and fixes the opening lines only.