A hook pack is six openings, one brief
create_hook_pack plans six distinct openings from one brand brief and charges once. Six prompts is not a pack.
A hook pack is six openings against one offer, planned so they do not repeat, charged once for the set. It is not six prompts you wrote, not six full ads, and not a library of generic lines. Create a branded hook pack is create_hook_pack: you brief the brand in a sentence or two, the agent plans N distinct hooks, quotes one charge, and submits every clip as a 1080p portrait job on Vidu Q3 with the line burned on as large TikTok-style text. Six is the test-design default when you actually intend to test. One brief. Six jobs. One bill.
This is not a UGC copy library. Category-specific lines already have a page. This is the object: width six, source one.
One brief, not six prompts
Required input is brand_context. Count, caption mode, vibes, emotions are optional. The planner’s job is distinctness. If you paste the six lines you already like, you have turned the pack into a renderer and thrown away the only reason it is a pack. Brief the offer. Let it propose the lines.
A good brief names:
- The product and the claim you will actually make.
- The audience and one objection.
- Words you will not claim.
- Whether the burned caption is the planned line or your copy.
A bad brief is “make it viral” with no SKU. The planner will invent six cousins of a joke. That set is spent.
Six separate generate_videos calls are six prices and, usually, six cousins of one sentence, because the calls cannot see each other. Pack billing inverts that. The cost is shown before you confirm. It does not climb per hook the way six independent video jobs would. The marginal hook inside the pack is the one you would not have written.
Premade catalog hooks (list_hooks) are a different SKU: existing reaction clips, no brand planning. Marketplace faces are a third SKU. Do not pay pack price to browse a face. Do not expect the pack to be a face library.
The UGC video generator is a related launcher for a body. The pack is openings. UGC is the look. Neither tool is a movie maker.
Why six, not one hero
Three openings are a read: you can see whether the planner understood the offer. Five or more is a test. Six is a practical test width — enough jobs to cover a proof-first open, a callout, an objection, a price lead, a sensory proof, a pattern interrupt — without pretending you ran a 20-cell matrix.
Ads with UGC get roughly 4× click-through and about 50% lower CPC, and 86% of brands say UGC lifts performance (Kapwing, 2026 UGC statistics; Billo). Those lifts assume the first seconds are a peer, not a brand wearing a hoodie. One hero line you walked in with was never tested against the other five jobs. Pack billing is why you can afford to find that out.
Width is a test-design choice, not a shopping-cart quantity. Do not pick three because you are picturing per-hook video pricing. Pick three only if three jobs cover the brief. Do not pick twelve because twelve sounds like science. Twelve near-duplicates is a planner miss you will still have to score.
Vidu Q3 is fixed on purpose. You are not picking Veo versus Seedance inside the pack. Comparability is the product. Route a hero film somewhere else after you pick the line.
Each finished hook comes back with the line on screen so you can score them the way a muted feed will see them. That burned line is the hook, not a transcript of a VO you will add later. If the body will speak a different sentence, you are testing the wrong object.
Distinctness is the product
When the pack returns, score distinctness first. If two clips are the same job in different adjectives, the planner missed. Do not “pick the nicer one” and call it a test. You bought coverage. Use it.
Jobs that should disagree:
- Proof first (the result, then the product).
- Objection first (the failure mode last Tuesday).
- Price or offer first.
- Sensory or demo first (the texture, the click, the pour).
- Pattern interrupt (a visual that is not a talking head, if the brief allows it).
- Callout (who this is not for).
If all six are “this changed my routine” with a different adjective, you do not have six openings. You have one opening at six-credit-looking width. Say so. Run another pack with a tighter brief — a second set, not a refund of the first.
Do not attach a generic UGC hook library to this object and call it planning. A library is copy you might type. A pack is clips the planner made from your SKU so they do not repeat. A UGC hook library that is specific is lines. This page is six files.
Human authenticity still wins surveys against AI UGC (81% versus 63% in the same UGC reporting). A pack does not erase that. It gives you six first seconds to test. If the winner still needs a real creator, you learned which line to hand them.
Attach winners to a body you already trust
The pack is openings, not a film. Billing once for the set is a reason not to regenerate the body six times to “match the energy” of each open. Lock a body — a demo, a talking take, a product still you would ship. Cut the winning open in front. Caption the body if it speaks; the pack already burned its own line.
Do not run Movie mode to “finish” a hook pack. Multi-scene films are a different production shape. Do not run six lipsync avatars because the pack felt cheap. The pack is Vidu Q3 portrait opens. An avatar is a mouth pass on a plate.
Article 50 still applies if the openings are realistic synthetic people and you publish them as authentic-looking media in the EU. Label the ads. Article 50 disclosure is the copy. A burned hook line is not the AI label.
The billing-shape post is the agent hook pack charges once for the set. The “not a film” post is create_hook_pack is openings, not a film. This page is only width six from one brief.
If you need a set of distinct openings for one offer, brief once, count six, confirm the one charge. Sign up and buy the set, not six cousins.
FAQ
Does a six-hook pack cost six video jobs?
No. The pack is charged as a set, shown before you confirm. Width is a test-design choice. Six independent generates are six prices and usually six cousins.
Can I name Veo inside the pack so the opens look more expensive?
Not as the point of this row. Comparability is the product. Vidu Q3, 1080p portrait, line burned on. Route a hero film on a named model after you pick the line.
What if I already have six lines I like?
Then you do not need the planner. Render those six as ordinary video jobs, or paste them only if you accept that you threw away distinctness planning. The pack exists because most people do not already have six jobs.