🎬 AI Video Ads Generator (Wan 2.2)

This actor generates AI-powered video ads for your products, brands, or services. You can create professional-quality videos for social media platforms like Instagram, TikTok, YouTube, and Facebook in minutes.

It supports 1–3 video variants per run, customizable according to your product, audience, ad goal, brand tone, and visual style. The actor uses the Wan2.2-I2V-A14B model for generating high-quality videos.

---

Features

• Generate multiple video ad variants in a single run.
• Generate high performing ads for your social media compaigns.
• Customize video format, resolution, and duration.
• Supports optional reference images for visual guidance.
• Logs progress during video generation so users know to be patient.
• Pushes full metadata to Apify dataset for tracking and billing.

---

What kind of data can be collected?

We do not collect any personal user data. All inputs, generated videos, and metadata belong to the user running the actor. Each user only has access to their own dataset.

---

💰 Why Choose This Actor? (Pay-for-Success)

Most AI video generators charge a high flat fee regardless of whether the AI "hallucinates" or fails to render. We believe you should only pay for cinematic value.

Zero Risk Entry: We charge a minimal "Handshake Fee" ($0.20) to initiate the GPU. If the API fails or times out, you aren't stuck with a full bill.

Dynamic Tiers: Using an advanced "Resource Ladder," our Actor calculates costs based on your specific settings. Want a quick 3-second preview? You pay significantly less than a full 10-second HD production.

High-Fidelity Assurance: The "Success Surcharge" only triggers when a valid video URL is generated and stored for you. Stop paying for AI "errors" and start paying for results.

---

💰 Estimated Run Costs

Configuration Type	Duration	Resolution	Estimated Total Price
Draft Preview	3 Seconds	Standard (SD)	~$3.50
Social Media	5 Seconds	High (HD)	~$3.80
Cinematic Master	10 Seconds	High (HD)	~$4.70

Note: These are estimates based on a successful run. You are only charged the bulk of the fee upon successful video generation.

Actor's Input

There are required and optional inputs.

Required Inputs:

Input	Type	Description
productName	string	Name of your product. Example: "Organic Green Tea"
referenceImageUrls	array	1–3 reference image URLs (for video style guidance). Example: ["https://example.com/img1.jpg"]
targetAudience	array	Target audience. Example: ["Women 18-35", "Fitness Enthusiasts"]
platform	string	Target platform for the ad. Options: "instagram", "tiktok", "youtube", "facebook"
videoFormat	string	Video format. Options: "auto", "portrait", "landscape_hd", "landscape", "square", "low"
adDuration	string	Duration of the ad in seconds. Options: "3", "5", "10"
basePrompt	string	Base prompt to guide AI video generation. Example: "Create an engaging ad for our organic green tea..."

Optional Inputs:

Input	Type	Description
adGoal	string	Goal of your ad. Example: "Increase brand awareness"
brandTone	string	Tone of your brand. Example: "Luxury", "Fun"
brandStory	string	Describe your brand story in 10–40 words. Example: "Our brand started with a passion for organic tea..."
visualStyle	string	Video style. Options: "Cinematic", "Animated", "UGC", "Artistic", "Normal"
keySellingPoints	string	Key product features, separated by commas. Example: "No Sugar, Organic, Vegan"
colors	array	Up to 3 hex colors to match brand identity. Example: ["#ff0000","#00ff00"]
variants	string	Video variants mode. Options: "none", "similar", "different"

---

Actor's Output

The actor generates video ad variants and pushes them to the Apify dataset. Each dataset row contains the following fields:

Field	Description
productName	Name of the product
referenceImageUrls	List of reference images
targetAudience	List of target audience segments
adGoal	Goal of the ad
platform	Target platform
videoFormat	Selected video format
width / height	Resolution of the generated video
adDurationSeconds	Duration of the ad video in seconds
brandTone	Brand tone
brandStory	Brand story text
visualStyle	Selected visual style
keySellingPoints	Comma-separated key selling points
colors	Brand colors
basePrompt	Base prompt used for AI generation
variantsMode	Selected variant mode (none, similar, different)
variant	Variant index (1, 2, or 3)
originalVideoUrl	Together AI-hosted video URL
apifyVideoUrl	Apify KV-hosted video URL
modelName	Model used (Wan-AI/Wan2.2-I2V-A14B)
createdAt	Timestamp of video creation

---

Example Inputs

{
  "productName": "Organic Green Tea",
  "referenceImageUrls": ["https://example.com/img1.jpg", "https://example.com/img2.jpg"],
  "targetAudience": ["Women 18-35", "Fitness Enthusiasts"],
  "adGoal": "Increase brand awareness",
  "platform": "instagram",
  "videoFormat": "portrait",
  "adDuration": "5",
  "brandTone": "Luxury",
  "brandStory": "Our brand started with a passion for organic tea and a mission to provide healthy beverages.",
  "visualStyle": "Cinematic",
  "keySellingPoints": "No Sugar, Organic, Vegan",
  "colors": ["#00ff00", "#ffcc00"],
  "basePrompt": "Create an engaging ad video for our organic green tea.",
  "variants": "similar"
}

🚀 Prompting Guide

Better prompts lead to cinematic results.

If you want to generate short video of your product from reference image, then it is good to provide only one clear reference image. Image should show your product clearly with white background and good lightning. Image urls must be publicly accessible and direct links to image files (e.g., ending in .jpg, .png). In base prompt, you can write "Use the reference image as the exact base frame. Do not change the person’s face, body, clothing, or product. Preserve identity, pose, background, and proportions. Do not hallucinate, etc." Please note that better prompting leads to better results.

For longer videos, use 3 high resolution reference images with different camera angles. You can write your brand story, Ad goal and other inputs. You can choose "similar" or "different" variants. For better results, try different inputs to check, which one gives you best results.

Retention policy

The retention policy of data for this actor is 31 days. However, there is also a video url generated by third-party service and they wrote that url link is valid for 7 days. However, links can be broken for several reasons before this. Therefore, this actor also downloads and save video in Apify store. You can access that link through "apifyVideoUrl" output. Still it is best for you you download a video on local device and export data.

Error Handling & Support

The actor can fail due to many reasons because it rely on third party services. We tested it and always trying to improve it's performance. It is good to wait for actor to complete it's job. However, if you found any technical issue or bug you can create an issue on the Actor'sIssues tab.

💸 Payment Structure

This Actor is currently transitioning to a pay-per-event monetization model to offer you more flexibility and lower entry costs. During the initial 14-day publication period, it uses a pay-per-result (PPR) model at approximately $3.50 per successful video. Once the transition is complete, the pay-per-event model will take over, allowing for dynamic pricing where standard generations start as low as $3.00, while premium high-definition and long-duration videos are capped around $4.70.

✨ What's Next?

Currently, this Actor specializes in cinematic visual generation and does not support audio. We are hard at work on our next major update, which will introduce another actor that works with it. This update will also feature synchronized background music and AI-generated voiceovers to complete your marketing toolkit.

License

This actor is provided as-is. The user who generates content using the Wan2.2 AI models owns the resulting content. Alibaba, the developer of the Wan2.2 series, explicitly states in the license agreements that they "claim no rights over your generated contents. You are fully accountable for your use of the models, which must not involve sharing any content that violates applicable laws, causes harm to individuals or groups, disseminates personal information intended for harm, spreads misinformation, or targets vulnerable populations.