Clipcat Prompt Library: 17,160+ Real-GMV Winning Prompts
Not a generic AI prompt directory. A directory of TikTok selling-video prompts that already produced GMV. 15 countries × 30 categories × 13 formats. Reusable in one click.
What the library is
The Clipcat Prompt Library is a curated set of 17,160+ TikTok selling-video prompts, each backed by a real TikTok Shop winner that produced verified GMV.
Most "AI video prompt directories" online are brainstormed collections. The Clipcat library is a spoils-of-war collection — the only inclusion rule is "this video actually sold something."
Why GMV-anchored, not view-anchored
In the TikTok ecosystem, high-view videos don't automatically sell. What actually drives sales is the combination of completion rate, engagement, add-to-cart and order rate — plus how well the video structure pushes a purchase decision.
- View-anchored: samples are "what people love to watch"
- GMV-anchored: samples are "what people watch and then buy"
For TikTok Shop sellers, only the second matters. That is the deepest distinction between Clipcat's library and public prompt directories.
Three filter axes
The library isn't a linear long list — it's a three-axis grid:
| Axis | Enum size | Typical picks |
|---|---|---|
| Country | 15 | US / SG / BR / MY / JP / ID / PH / VN / GB / TH / IT / DE / ES / FR / MX |
| Category | 30 | Beauty, home, pet, fitness, baby, apparel, 3C, kitchen, auto... |
| Video format | 13 | Real usage review, spokesperson, hand-held demo, OOTD, lifestyle, promo push... |
Crossing the axes yields tight candidate pools like "US · beauty · real usage review" or "Indonesia · home · hand-held demo."
13 winning formats explained
The 13 formats aren't a design — they emerged from clustering 17,160+ real winners. Top six by sample count:
| Format | Samples | Core anatomy |
|---|---|---|
| Hand-held product demo | 7,189 | Hands + close-up + minimal callouts — the dominant form |
| Real usage review | 2,338 | Unbox → try → verdict, full decision chain |
| Promo hard push | 2,058 | Price / discount / urgency, three beats |
| Lifestyle scene | 1,450 | Product embedded in real life, emotion-led |
| OOTD | 1,119 | Apparel & beauty strength, string multiple SKUs into one outfit |
| Spokesperson | 993 | Talking to camera + subtitles — avatar-friendly |
Which format fits your category has already been answered by the library — via sample density.
15 country market distribution
Winner structure varies wildly by market — the US is content-depth-first, Southeast Asia is emotion-forward, Japan is texture-obsessed. Top eight by sample density:
- US · 5,488
- SG · 1,078
- BR · 992
- MY · 973
- JP · 967
- MX · 943
- ID · 916
- PH · 899
For US you can filter down to "beauty · lipstick · hand-held demo" granularity. For smaller-language markets, start on "category + format" first and layer country on top.
Combining with one-click clone
The library is a node in Clipcat's full loop, not a standalone tool. Standard usage:
- Filter candidates by country + category + format
- Browse the shortlist, pick 1-3 winners that best fit your SKU
- Load the prompt into the clone panel, swap product name / hero shot / selling point
- One-click generate → one-click publish to TikTok
- Return 3-7 days later, tag winning structures into your "personal winner library" and keep reusing
The library's value isn't "how many prompts." It's "can you consistently pull one out each day that converts to GMV."
FAQ
Is the library priced separately?
Free tier includes browsing and using base samples. Paid tiers unlock niche categories, non-mainstream markets, deep filters and batch export.
Can I export prompts to another tool?
The library is designed to work with Clipcat's clone engine. Exported prompts alone won't reproduce the same winner structure elsewhere — winner structure = prompt + clone engine + sample reference, and all three are needed.
How often is it updated?
Rolling updates as TikTok Shop winners get GMV-anchored — daily additions.
Does the library support English?
Yes. Bilingual — you can switch between English and Chinese views.
How do I judge if a prompt fits my SKU?
Three checks: same category, same price band, same target market. The more aligned, the higher the clone success rate.