How it works
From one sentence to a finished UGC video ad
VydioFlow turns a prompt or a product URL into a structured campaign brief, writes the hooks and scripts, renders each scene as an individual AI video job, then hands you captions, branding and every export ratio.
How it works
From a sentence to an export-ready ad, in six steps
A 20s TikTok ad for my vitamin C serum — upbeat creator, bathroom mirror, price hook
Step 01 — Brief
Skip the shoot. Start with a sentence.
No casting calls, no shipping product to creators, no waiting three weeks for a first cut. Describe the ad or paste your product URL and VydioFlow builds the brief, the angle and the offer language in seconds — then shows you the exact credit cost before anything renders.
- Prompt, product URL, or your own finished script
- Offer, audience and objections extracted automatically
- Cost shown up front — nothing renders without your go-ahead
9:16 · 9s · captions on
Step 02 — Render
Scenes render like a real edit suite
Each scene is its own job with its own prompt, so the cut behaves like a timeline instead of a slot machine. Reroll the scene that missed, keep the two that landed, and ship the ad the same afternoon it was briefed.
- Scene-level regeneration — pay only for what you replace
- Renders continue if you close the tab
- Captions, brand colours and CTA end screen applied automatically
9:16 · 8s · captions on
What happens on our servers
Every model call runs server-side. Script, hook and prompt writing use Google Gemini; scene video renders use Google Veo. API keys never touch browser code, and each render is stored as a generation job with its model, prompt, duration, status, cost and output URL so you can audit exactly what was produced and what it cost.
Scenes are independent jobs. If scene three misses, you rewrite that prompt and re-render it alone — the rest of the cut stays untouched, and you only spend credits on the scene you replaced.
What you get back from a single run
Multiple ad concepts
Distinct angles for the same offer so you can test messaging, not just visuals.
Hooks and full scripts
Opening lines written to stop the scroll, plus the scene-by-scene script beneath them.
Rendered scenes
AI UGC footage with a consistent creator across the cut, rendered asynchronously with live status.
Captions
Burned-in style captions timed to the script for sound-off feeds.
Brand kit applied
Your colours, logo and CTA end screen reused automatically on every export.
Every ratio
9:16, 1:1 and 16:9 from the same project, without re-editing per channel.
4 steps
From a prompt or product URL to an exported ad
~3 scenes
Typical finished ad, each rendered as its own job
3 ratios
9:16, 1:1 and 16:9 exported from one project
1 credit
Cost of a full hook and script pass before any render
A closer look at each step
The same pipeline runs whether you start from a sentence, a pasted script or a product page.
- 01
Brief the offer
Type what the ad should sell or paste a product URL. The page is read and turned into a structured brief — product, audience, promise, objections and the CTA you want at the end.
- 02
Pick an angle and hook
Several distinct concepts are written for the same offer. Choose the angle that fits the channel, then swap between hook variations until the opening line earns the first three seconds.
- 03
Edit the script before it costs anything
Scripts are fully editable text. Rewrite a line, shorten a scene, change the CTA — nothing renders and no scene credits are spent until you approve the cut.
- 04
Cast a creator and set quality
Choose an AI creator to keep one face across the cut, or go creator-free with product and lifestyle scenes. Pick Standard for drafts, Pro for campaigns, Cinema for hero spots — the exact credit cost is shown before you start.
- 05
Render scene by scene
Each scene is queued as an independent job with live status. Renders keep going if you close the tab, and a missed scene is re-rendered alone instead of paying for the whole ad again.
- 06
Caption, brand and export
Timed captions, your logo, colours and a CTA end screen are applied automatically from the brand kit, then exported in every ratio you need for TikTok, Reels, Shorts and YouTube.
Common questions about the workflow
- How long does a render take?
- A single scene usually completes in a couple of minutes; a three-scene ad typically finishes in under ten. Scenes render in parallel and you can leave the page — jobs are reconciled when you come back.
- Do I have to use an AI creator?
- No. Creators are optional. Plenty of ads perform better as product-led b-roll, screen recordings or lifestyle cutaways with voiceover and captions.
- Can I bring my own product photos?
- Yes. Upload product shots, screenshots or a logo and they are used as reference frames, so the product on screen is genuinely yours rather than an invented lookalike.
- What if a scene comes out wrong?
- Rewrite that scene's prompt and re-render it on its own. Only the replaced scene costs credits; the rest of the cut is untouched.
Ads that came out of this exact workflow
Every clip below started as one sentence in the same composer you get on day one.
Local services
Cinematic B-roll with bold kinetic headline
Fashion
Styling montage with kinetic titles
Mobile app
Product-in-hand demo with a soft-sell close
Run the pipeline on your own offer
Start with one product URL and see the hooks, script and first rendered scene before you commit a campaign to it.