AI Generation Overview — What You Can Create
Last updated: April 18, 2026
Wevion includes a built-in AI creative engine that lets you generate images, videos, avatar spokesperson clips, voiceovers, and fully assembled composite videos — all without leaving the platform. Instead of juggling separate tools and file transfers, you produce ad creatives directly inside Creative Hub and use them in campaigns immediately.
This article covers what you can generate, how the process works, and where to find detailed guides on each content type.
Prerequisites
Before using AI Generation, make sure you have the following:
- An active Wevion account with a valid subscription that includes AI credits
- At least Media Buyer role or higher in your team
- Access to Creative Hub (visible in the main sidebar navigation)
No additional software, accounts, or API keys are required on your end. Wevion manages all AI infrastructure behind the scenes.
The AI Generation Panel
You access AI Generation through the magic wand icon in Creative Hub. Clicking it opens the Generation Panel — a slide-out interface where you configure and submit generation jobs.
The panel is organized by content type. Select what you want to create, fill in the required fields (at minimum, a text prompt or script), adjust optional settings, and hit Generate. Your job enters the queue, and the finished asset appears in your Creative Hub folder when it is done. You can continue working while generation runs in the background.
What You Can Create
Wevion supports five types of AI-generated content. Each is powered by multiple AI engines selected automatically, or you can choose a specific engine.
Images
Generate still images from a text prompt — product shots, lifestyle scenes, abstract backgrounds, anything you describe.
- Text-to-image: Describe your vision in natural language
- Image-to-image: Provide a reference image to generate variations
- Multiple variants: Request several options per generation
- Ad-ready formats: Generate at Meta's exact dimensions — no resizing needed
Supported formats:
| Format | Dimensions | Best for |
|---|---|---|
| Feed Square (1:1) | 1080 x 1080 | Facebook/Instagram feed ads |
| Feed Portrait (4:5) | 1080 x 1350 | Instagram feed, highest engagement ratio |
| Stories/Reels (9:16) | 1080 x 1920 | Stories, Reels, vertical placements |
| Landscape (16:9) | 1920 x 1080 | In-stream video covers, desktop placements |
| Right Column | 1200 x 628 | Right column, link preview images |
Videos
Generate short video clips from a text prompt, or animate a still image using image-to-video.
- Text-to-video: Describe a scene and receive a video clip
- Image-to-video: Turn product photos or AI images into dynamic video ads
- 2-10 second clips with adjustable duration
- Aspect ratios: 16:9 (landscape), 9:16 (vertical/Stories), or 1:1 (square)
Expect 30 seconds to a few minutes depending on duration and engine.
AI Avatars (UGC-Style Videos)
Create spokesperson videos where a digital avatar delivers your script — the talking-head UGC format that performs well in direct-response ads.
- Script-driven: Up to 5,000 characters with natural lip sync
- Avatar library: Browse and choose a persona that fits your campaign
- Voice selection: Multiple voices with different tones and styles
- Multi-language: Generate in different languages for international campaigns
Avatars are the most resource-intensive type, taking several minutes depending on script length.
Creating AI Avatars (UGC Videos)
Text-to-Speech (TTS)
Generate natural-sounding voiceover audio from text. Use these files as narration in composite videos or download them for external editing.
- Natural voices: Multiple AI voice engines produce realistic, human-sounding speech
- Voice library: Choose from voices with different characteristics — gender, accent, tone
- Multi-language: Generate speech in multiple languages
- Up to 5,000 characters per generation, covering narration for ads of any length
TTS is the fastest content type — audio typically generates in a few seconds.
Text-to-Speech (TTS) — Voice Generation
Composite Videos
Combine video clips, images, audio tracks, and text overlays into a single finished creative.
- Multi-layer: Mix videos, images, audio, and text in one output
- Templates or custom: Pre-built templates or fully custom layouts
- Precise timing: Control when each element appears and disappears
- Positioning: Place elements at specific coordinates with styling
- Flexible resolution: Standard ad sizes or custom dimensions
Typical workflow: generate or upload individual assets first, then combine them. Swap a voiceover or background without regenerating everything.
Video Compositing — Assembling Multi-Element Videos
How Generation Works
Every generation request follows the same lifecycle:
- Submit — Fill in the form and click Generate
- Pending — Wevion creates a job record and queues it for processing
- Processing — The AI engine works on your content
- Completed — The finished asset is saved directly to your Creative Hub folder
- Ready to use — The file is immediately available for preview, download, or campaign use
If something goes wrong — a content safety filter triggers or a temporary issue occurs — the job status changes to Failed with an error message. Failed jobs do not consume credits.
Destination folder: By default, assets save to your root Creative Hub folder. You can specify a destination folder before submitting so the output lands exactly where you want it.
Checking status: Monitor all generation jobs from the Jobs panel. Filter by type (image, video, avatar, TTS, composite) or status (pending, processing, completed, failed).
Managing AI Jobs — Status and Download
Credits System
AI generation uses a credits system tied to your subscription plan. Each generation consumes credits based on content type, duration, and engine.
- Credits deducted on completion, not on submission
- Failed generations do not consume credits
- Costs vary by type — TTS is inexpensive, avatar videos cost more
- Balance visible in the Generation Panel and account settings
Getting Started
- Open Creative Hub from the main sidebar navigation
- Click the magic wand icon to open the AI Generation panel
- Select Image as your first content type — it is the fastest to generate
- Enter a prompt — be specific: "A flat-lay product photo of a fitness tracker on a marble surface, natural lighting, minimal style" produces better results than "product photo"
- Choose a format — select the dimension that matches your ad placement
- Click Generate and wait for the result to appear in your folder
- Preview the result — click the file to see it at full resolution
From there, experiment: animate your image with image-to-video, add a TTS voiceover, then combine everything into a composite video.
FAQ
How long does generation take? Images: 10-60 seconds. Videos: 30 seconds to a few minutes. TTS: a few seconds. Avatars: several minutes. Composites: varies by complexity.
Can I cancel a job? Once a job enters Processing, it cannot be cancelled. Let it complete and discard the result if needed.
What if I run out of credits? Credits are checked before submission. If you do not have enough, the request will not be accepted.
Are there content restrictions? Yes. AI engines enforce content safety policies. Prohibited content prompts will be rejected with an error message.
Can I regenerate with the same prompt? Yes. Generation is non-deterministic, so the same prompt produces different results each time.
Do I need to download files before using them in campaigns? No. Generated files save directly to Creative Hub, and Campaign Launcher pulls from there.
Which AI engines are used? Wevion integrates multiple engines per content type for quality and reliability. The platform selects the best engine by default, or you can choose a specific one. You interact with results, not infrastructure.