For

Vyra for UGC creators: hook-and-demo ads from raw takes, in the brand's style

Vyra turns your raw takes into hook-and-demo ads, testimonials, and unboxings. It picks the best take of each line and matches the caption style a brand wants.

By Sulan Zhang, co-founder of Vyra · Updated 2026-09-24

Best for: UGC creators delivering paid ads and organic-style content to brands
Typical footage: Phone takes of the hook, demo, and close, product b-roll, screen recordings
Typical output: 15-45 second hook-and-demo ads, 3-5 variants per brief, testimonials, unboxings
Time saved: One shoot to five ad variants with different hooks in one conversation
Works with: Built-in chat, Claude, ChatGPT, Cursor over MCP
Price: Free to start, plans from $24/mo

UGC work is volume, variants, and brand specs. Vyra picks the cleanest take of each line, builds the hook-demo-close structure from a sentence, and copies the caption style a brand sends you as a reference.

A real week with Vyra

A skincare brand sent a brief on Monday. 30 seconds, three hook options, word-by-word captions in their font, logo end card. She filmed 11 takes of the hook, one demo, and one close.

Build a 30-second 9:16 hook-and-demo ad. Demo is my take where I apply it and name the three ingredients, cut to the product close-up each time I name one. Close is the take where I say the brand name. Then make three versions that differ only in the hook. Use my three cleanest hook takes. Captions word-by-word, Montserrat Black, white, and a 3-second logo end card.

Three variants came back with the same demo and close and different first 4 seconds. The brand's reference video went in as a second prompt, "match this caption color and highlight," and all three updated. Delivered Wednesday, raw takes included from the library. The brand picked hook 2.

What you make

Your three most-used prompts

Hook-and-demo

Build a 30-second 9:16 hook-and-demo ad. Hook is my cleanest take of "I stopped buying these." Demo is the product b-roll while I explain the three reasons, cut to the product each time I name one. Close is the take where I hold it up and say the brand name. Word-by-word captions, bold, centered high. No music.

Variants

Make four more versions of that ad, each with a different hook take from the footage. Keep the demo and close identical. Name the exports hook-1 through hook-4.

Brand caption match

Match the caption style from the brand's reference video. Font weight, color, position, and the highlight color on the active word. Apply it to all five variants.

A typical workflow

Upload every take and the product b-roll.
Describe the structure once. Ask for the best take of each line.
Ask for hook variants from the same footage.
Attach the brand's reference and match the captions.
Export all variants at 9:16 and deliver.

What Vyra does that matters for you

Take selection by transcript. "The cleanest take where I say X."
Variants without re-editing. Swap one segment, keep the rest.
Reference style matching for brand caption specs.
Consistent output across a batch, which is what brands pay for.

What it does not do

It doesn't generate AI avatars or synthetic UGC. You're the creator on screen.
It doesn't run the ads or report performance.

Example

Example creator: @sulansart

FAQ

Can I deliver raw plus edited? Yes. Export the cut. The original uploads stay in your library.

How do I hit a brand's exact spec sheet? Paste the spec into the prompt. Length, aspect, caption rules, logo placement, safe zones. It follows written rules well.

Can I reuse a structure for the next brand? Yes. Say "same structure as the skincare ad, new footage."