Gentic Creative Assets — Documentation
Everything you need to generate brand ad creatives, search inspiration ads from top brands, and manage your product image library.
1. Getting Started
Sign Up & Get Your API Key
Before you can use Gentic Creative Assets, you need an API key to authenticate your requests.
- Go to gentic.co/creative and create an account.
- Create an organization from your dashboard. API keys, billing, and asset data are all scoped to the organization.
- Go to the API Keys section in your dashboard.
- Click Create API Key. Give it a name (e.g. "Claude Code" or "n8n production").
- Copy the key immediately — it starts with
gentic_and is only shown once.
Anyone on your team can generate their own key, and they'll all share the same asset library and credit balance.
Keep your key secure. Treat it like a password. Don't commit it to version control or share it publicly.
2. Connecting to the MCP Server
Gentic Creative Assets uses the Model Context Protocol (MCP) — an open standard for connecting AI agents to external tools. You connect once, and your agent gets access to all Gentic Creative Assets tools automatically.
Server URL: https://mcp.gentic.co/creative
Claude Code (CLI)
Option A — OAuth (no API key needed):
claude mcp add gentic-creative --transport http https://mcp.gentic.co/creativeClaude Code will prompt you to authenticate on first use.
Option B — API key:
claude mcp add gentic-creative --transport http https://mcp.gentic.co/creative --header "Authorization: Bearer <your-api-key>"Replace <your-api-key> with a key from your dashboard.
Claude Web / Claude Desktop / Claude Mobile
- Go to Settings → Connectors
- Click Add custom connector at the bottom of the page
- Enter the server URL:
https://mcp.gentic.co/creative - Click Add — you'll be redirected to sign in and authorize access via OAuth
Custom connectors require a Claude Pro, Max, Team, or Enterprise plan.
ChatGPT
- Go to Settings → Apps & Connectors → Advanced settings and enable developer mode
- Go to Settings → Connectors → Create
- Set the connector name to Gentic Creative Assets and the URL to:
https://mcp.gentic.co/creative - You'll see Gentic's tools listed if the connection is successful
- In a new chat, click + near the message bar, select More, and choose Gentic Creative Assets
ChatGPT Apps are available on Plus, Pro, Team, Business, Enterprise, and Education plans.
n8n
- Add an MCP Client node to your workflow
- Set the server URL to
https://mcp.gentic.co/creative - Add your API key as a Bearer token in the authentication settings
Any Other MCP Client
Gentic Creative Assets uses standard Streamable HTTP transport. Any client that supports MCP can connect using the server URL and either OAuth or API key Bearer token authentication.
Agent Skills (Recommended)
For the best results, pair the MCP server with the Gentic agent skill. The MCP server gives your agent access to the tools; the skill teaches it the optimal workflow order — catalog setup, image search, ad generation with inspiration ads, and job tracking.
1. Add the MCP server (if you haven't already):
claude mcp add gentic-creative --transport http https://mcp.gentic.co/creative --header "Authorization: Bearer <your-api-key>"2. Install the agent skill:
npx skills add gentic-co/agent-skills- MCP server — provides tool access (generate, vectorize, search, etc.)
- Agent skill — teaches the optimal workflow order: catalog setup → image search → ad generation → job tracking
Works with Claude Code, Cursor, Copilot, and 40+ other agents.
3. Vectorize Product Images
Tool: vectorize_product_images
Cost: $0.05 per image
Before you can search your product images by description, they need to be vectorized. This tool takes image URLs and creates multimodal embeddings that combine the visual content of each image with any text context you provide — enabling powerful natural language search later.
Parameters
| Parameter | Required | Default | Description |
|---|---|---|---|
| images | Yes | — | Array of image objects, each with an image_url (required) and optional product_name |
Image object format
| Field | Required | Description |
|---|---|---|
| image_url | Yes | Public URL to the product image |
| product_name | No | Name of the product (e.g. "Ocean Blue Sneakers"). If omitted, AI vision will auto-generate a name from the image. |
Key behaviors
- Batch processing: 1–50 images per call
- Async — the tool returns immediately and processing happens in the background
- Uses multimodal embeddings that understand both image pixels and text semantics
- Images are stored in your org's vector index for search
Example prompts
Basic vectorization:
"Vectorize these product images: [url1], [url2], [url3]"
With product names:
"Vectorize these product images. The first is our Vitamin C Serum and the second is our Overnight Recovery Mask."
Pro tip
Always include product names when vectorizing — they significantly improve search accuracy. If you omit the name, the AI will auto-generate one from the image, but explicit names are more reliable. Use fetch_page ($0.10) to pull product details from your website automatically.
4. Search Product Images
Tool: search_product_images
Cost: Free
Search your vectorized product image library using natural language. Describe what you're looking for and the tool returns the most relevant images from your catalog — useful for finding the right product shots to reference when generating ad creatives.
Parameters
| Parameter | Required | Default | Description |
|---|---|---|---|
| search_text | Yes | — | Describe the kind of product image you're looking for (e.g. "blue athletic sneakers", "luxury gold watch") |
| limit | No | 5 | Number of results to return (1–50) |
Example prompts
By product type:
"Find product images of our sunscreen products."
By visual style:
"Search for product photos with a white background and minimal styling."
By use case:
"Find images that would work well for a summer campaign — bright colors, outdoor feel."
How search works
Your text query is converted into a multimodal embedding and compared against your vectorized image library. Because the embeddings capture both visual and semantic information, you can search by product type, visual style, color palette, or abstract concepts — not just keywords.
Results include the original image URL, any text context provided during vectorization, and a similarity score.
5. Search Inspiration Ads
Tool: search_inspiration_ads
Cost: Free
Instead of manually browsing the Meta Ads Library for creative references, let your agent search a curated database of 20,000+ vectorized ads from the top 1,000 ecommerce brands. Only the highest-impression ads are indexed — so every result is a proven, high-performing creative. Use the returned image URLs as inspiration_image_urls in generate_ad_asset to create ads inspired by a competitor's style.
Parameters
| Parameter | Required | Default | Description |
|---|---|---|---|
| search_text | Yes | — | Describe the kind of ad you're looking for (e.g. "minimalist skincare ad with white background", "bold fitness ad with athlete") |
| brand_name | No | — | Filter to a specific brand (e.g. "Clean Skin Club", "Nike") |
| category | No | — | Filter by industry category (e.g. "Health/beauty", "Apparel") |
| platform | No | — | Filter by platform: "FACEBOOK" or "INSTAGRAM" |
| active_only | No | true | Only return currently active ads |
| limit | No | 10 | Number of results to return (1–50) |
Example prompts
By visual style:
"Find minimalist skincare ads with clean white backgrounds and soft lighting."
By competitor:
"Show me the top-performing ads from Glossier."
By category and platform:
"Find fitness apparel ads running on Instagram with bold typography and action shots."
How it works
Gentic has vectorized 20,000+ ads sourced from the Meta Ad Library, focusing exclusively on the top 1,000 ecommerce brands with the highest impression counts. Your text query is converted into a vector embedding and matched against this curated database — so every result is a real, high-performing ad from a proven brand.
This replaces the manual workflow of browsing the Meta Ads Library to find creative inspiration. Instead of spending hours searching, filtering, and screenshotting competitor ads, your agent can surface the most relevant references in seconds and pipe them directly into ad generation.
Pro tip
Combine search_inspiration_ads with generate_ad_asset for a fully automated creative workflow: search for competitor ads by style, then pass the returned URLs as inspiration_image_urls to generate ads that match the visual quality of top-performing brands — but with your own products and branding.
6. Generate Ad Creatives
Tool: generate_ad_asset
Cost: $1.00 per image generated
This is the core tool. Describe the ad creative you want and Gentic generates it using AI. Provide inspiration ads from competitors to emulate their visual style, add your own brand images (product photos, logos), and configure aspect ratios and resolutions up to 4K. Batch generate up to 10 images at once.
Parameters
| Parameter | Required | Default | Description |
|---|---|---|---|
| prompt | Yes | — | Detailed text prompt describing the ad image to generate |
| inspiration_image_urls | No | [] | 0–5 reference/competitor ad URLs to use as style inspiration. The AI will emulate the visual style, layout, and aesthetic — but NOT copy the content. |
| brand_image_urls | No | [] | 0–5 brand asset URLs (product photos, logos) to incorporate into the generated image |
| count | No | 1 | Number of images to generate in parallel (1–10). Use this instead of calling the tool multiple times. |
| aspect_ratio | No | "1:1" | Aspect ratio: "1:1", "2:3", "3:2", "3:4", "4:3", "4:5", "5:4", "9:16", "16:9", "21:9" |
| image_size | No | "4K" | Output resolution: "1K" (1024px), "2K" (2048px), or "4K" (4096px) |
Example prompts
Simple generation:
"Generate a product lifestyle image for our Vitamin C Serum — warm morning light, bathroom counter, minimal and clean."
With inspiration ads:
"Generate an ad creative for our SPF launch. Use these competitor ads as style inspiration: [url1], [url2]. Use our product photo as the brand image: [url3]. Make it 9:16 for Instagram Stories."
Batch for A/B testing:
"Generate 5 variations of a hero banner for our SPF launch. 16:9 ratio. Mix lifestyle and product-focused angles."
How generation works
Generation is asynchronous. When you trigger it, Gentic sends the job to its AI image generation engine. The tool returns immediately with job ID(s) — use list_asset_jobs to check progress and retrieve the output image URLs once completed.
Inspiration ads vs. brand images
- inspiration_image_urls — competitor or reference ads whose visual style, layout, and aesthetic the AI should emulate. The content is NOT copied — only the style is transferred.
- brand_image_urls — your own product photos, logos, or brand assets that the AI will incorporate directly into the generated image.
Pro tip
Be specific in your prompts. Instead of "product ad", describe the scene, lighting, mood, and composition. The more detail you provide, the closer the output will match your vision. Use search_product_images to find the right product shots first, then pass their URLs as brand_image_urls. Use search_inspiration_ads to find high-performing competitor ads, then pass them as inspiration_image_urls to match their visual quality.
7. Track Generation Jobs
Tool: list_asset_jobs
Cost: Free
Since ad generation is asynchronous, use this tool to check the status of your jobs, view completed output URLs, and monitor any failures.
Parameters
| Parameter | Required | Default | Description |
|---|---|---|---|
| limit | No | 20 | Number of recent jobs to return (1–100) |
| status | No | All | Filter by "processing", "completed", or "failed" |
What you get back (per job)
- Job ID — unique identifier
- Status — processing, completed, or failed
- Prompt — the description used to generate
- Image URLs — output image URLs (when completed)
- Timestamps — when the job was created and completed
Example prompts
Check all jobs:
"Show me my recent asset generation jobs."
Check completed only:
"Show me my completed ad creatives."
Check for failures:
"Did any of my generation jobs fail?"
8. Tool Reference
| Tool | Description | Cost |
|---|---|---|
| generate_ad_asset | AI-generate brand ad images. Supports batch generation (1–10), inspiration ads for style emulation, brand image references (up to 5 each), 10 aspect ratios, and resolutions up to 4K. | $1.00/image |
| list_asset_jobs | List recent asset generation jobs with status, prompts, and output image URLs. Filter by processing, completed, or failed. | Free |
| vectorize_product_images | Batch vectorize product images using multimodal embeddings. Combines image pixels and text context for semantic search. | $0.05/image |
| search_product_images | Search your vectorized product image library by natural language text description. | Free |
| search_inspiration_ads | Search 20,000+ vectorized ads from the top 1,000 ecom brands (highest-impression ads from Meta Ad Library). Filter by brand, category, or platform. | Free |
| fetch_page | Extract readable text from any web page. Useful for pulling product details or brand context. | $0.10/page |
9. Pricing
No subscriptions. No seat licenses. Pay only when your agent calls a paid tool.
| Action | Cost |
|---|---|
| Ad creative generation (per image) | $1.00 |
| Product image vectorization (per image) | $0.05 |
| Web page fetching | $0.10 |
| Asset job tracking | Free |
| Product image search | Free |
| Inspiration ad search | Free |
Example workflow cost
- Fetch product page: 1 x $0.10 = $0.10
- Vectorize 20 product images: 20 x $0.05 = $1.00
- Search product library: Free
- Search inspiration ads for style references: Free
- Generate 5 ad creatives at 4K: 5 x $1.00 = $5.00
- Track generation jobs: Free
- Total: $6.10
Credits are deducted from your organization's balance in real-time. You can check your balance anytime in your dashboard.