Chat with your creative · Brand-relative performance · MCP for any AI

Your creative
operating system

Creative Tagger is the context layer between your ad account and your LLM. Connect your creative and performance, then chat with it — or plug it into Claude, ChatGPT, or Cursor via MCP — to see what's working and decide what to make next.

Start Free Connect in 60 Seconds
Request
// POST /analyze
curl -X POST \
https://api.creativetagger.ai/analyze \
-F "file=@ad_video.mp4" \
-F "brand_name=Brand"
Response
{
"format": "video",
"attributes": {
"asset_type": "UGC",
"visual_format": "Talking Head",
"hook_type": "Question",
"talent": "Creator",
"audience": "Wellness Seekers",
"emotion": "Curiosity"
},
"brand_attributes": {
"founder": "Stephen Lavender",
"customer_segment": "High Intent Founder"
},
"recognized_entities": [
{ "type": "founder", "name": "Stephen Lavender" }
],
"naming": {
"standard": "BRAND_UGC_
TalkingHead_Creator_Question_ShopNow_9x16_V1"
}
}
Model Context Protocol

Plug into whatever AI you already use.

Creative Tagger ships as an MCP server. Claude Desktop, Cursor, Windsurf, ChatGPT — if your AI speaks MCP, it can call us. The full V1 toolset. One config block.

1. Install
# pip
pip install creative-tagger-mcp
2. Add to Claude Desktop
{
"mcpServers": {
"creative-tagger": {
"command": "creative-tagger-mcp",
"env": {
"CREATIVE_TAGGER_API_KEY":
"ct_your_key"
}
}
}
}
What your AI gets
analyze_creative — classify any creative across all dimensions
list_library — search past analyses (memory)
get_library_patterns — concentration + diversity insights
get_brand_taxonomy / set_brand_entity — founders, products, segments, aliases
recommend — strategist answers grounded in library + brand
analyze_gaps — what's missing + what to make next
get_brand_context / set_brand_context — long-term brand memory
scan_competitor / import_competitor_ads — classified Meta Ad Library breakdown, native or user-owned import
get_meta_status / sync_meta_performance — read-only performance memory
generate_naming — standardized file names from attributes
Creative Strategist

A strategist that remembers.

Every analysis is saved to your library. The strategist reads your library + your saved brand voice, audience, and anti-patterns — then answers strategic questions with concrete, taxonomy-backed briefs.

  • ✓Recall every creative you’ve ever tagged
  • ✓Get gap analysis: where you over-index, what to test next
  • ✓Recommend creatives using your brand voice and audience
  • ✓Runs inside your AI of choice — Claude, ChatGPT, Cursor
// Claude calls our MCP
recommend({
brand_name: "Acme",
question: "What should I test for Q4?"
})
// Strategist auto-loads memory
{
"answer": "You over-index on UGC...",
"recommendations": [{
"hook_type": "Curiosity Gap",
"messaging_angle": "Before After",
"creative_type": "Tutorial",
"script_outline": "..."
}],
"library_observations": [
"73% UGC concentration"
]
}

Every format. One endpoint.

Auto-detects format from your input. Same structured taxonomy across every creative type.

Video Ads

Short-form video with full visual, audio, and transcript analysis. Frame extraction, trending audio detection, AI transcription.

Hook Detection Audio Fingerprint Transcript

Static Images

Image ads with text overlay extraction, dominant color detection, and auto aspect ratio. Same visual taxonomy as video.

Text Overlay Color Analysis

Carousels

Per-slide classification plus aggregate analysis. Detects narrative arc and visual coherence across all cards.

Per-Slide Narrative Arc Coherence

Long-Form Video

YouTube and long-form content with automatic chapter detection, content structure analysis, and pacing pattern classification.

Chapters Structure Pacing

Landing Pages

Renders desktop and mobile via headless browser. Extracts headlines, CTAs, form complexity, social proof, and visual hierarchy.

Desktop + Mobile Social Proof

Email Creatives

HTML email analysis with subject line extraction, CTA clarity scoring, visual density classification, and personalization detection.

Subject Line CTA Clarity Personalization

Standard taxonomy. Brand-specific memory.

Creative Tagger keeps a stable base layer for benchmarking, then lets every brand extend it with allowed values, aliases, entities, segments, and naming rules. Your reports stay comparable; your strategist still speaks your brand's language.

Stable base layer
Same standard taxonomy every time. No more "UGC-style" vs "ugc" vs "User Generated".
Brand-custom extensions
Founder names, recurring creators, customers, products, offers, ICPs, segments, and internal campaign labels.
Naming that knows your terms
Templates can use standard dimensions plus brand-specific fields, so exports match how your team already works.
// standard taxonomy + brand layer
asset_type → UGC Lifestyle Product Shot
Studio High Production AI Generated
hook_type → Question Bold Claim Callout
Curiosity Gap Pain Point Contrarian
Transformation Pattern Interrupt
format → Talking Head Testimonial Demo
Before After Unboxing Founder Story
emotion → Urgency Curiosity Trust Desire
Humor Aspire Empower Nostalgic
// + audience, messaging angle, offer, audio, seasonality...
brand.founder → Stephen Lavender
brand.segment → High Intent Founder
brand.offer → Launch Sprint
// custom values never replace the standard `attributes` layer
Comparison

Motion locks the taxonomy.
Creative Tagger lets brands extend it.

Motion is powerful, but its intelligence still centers around Motion's dashboard and vocabulary. Creative Tagger gives you a standard taxonomy that each brand can customize without breaking reporting.

Motion App

Managed system

  • ×Analysis is centered around Motion's reporting workspace
  • ×Per-brand vocabulary is generated and managed inside their system
  • ×Founder, segment, and internal naming rules are harder to keep portable
  • ×MCP exists, but it is not an open brand-taxonomy layer you own
  • ×Seat/workspace economics can be heavier than usage-first teams need
Creative Tagger

MCP infrastructure

  • ✓Plugs into Claude, ChatGPT, Cursor — any AI you trust
  • ✓Use it from your AI chat, your code, your dashboard
  • ✓Standard taxonomy + Brand Taxonomy Studio for custom values and entities
  • ✓Library, brand memory, and Meta performance import exposed as MCP tools
  • ✓Pay per analysis — no seats, no minimums

Simple pricing

Free to start. Pay as you go. Dashboard and brand taxonomy available from day one.

Free
$0
50 credits/month
  • ✓ All 6 formats
  • ✓ No credit card required
  • ✓ MCP server + Python SDK
  • ✓ Dashboard + brand taxonomy
Start Free
Pay As You Go
$0.06/call
No subscription, billed monthly
  • ✓ Pay only for what you use
  • ✓ Auto-billed, no interruptions
  • ✓ Spending limit controls
  • ✓ API + Dashboard access
Add Payment
Most Popular
Starter
$49/mo
1,000 credits included
  • ✓ API + Dashboard access
  • ✓ $0.06/credit overage
  • ✓ Brand Taxonomy Studio
  • ✓ Meta performance memory
Start Starter
Growth
$149/mo
5,000 credits included
  • ✓ API + Dashboard access
  • ✓ $0.05/credit overage
  • ✓ Brand Taxonomy Studio
  • ✓ Meta performance memory
Start Growth
Scale
$499/mo
25,000 credits included
$0.04/credit overage
API + Dashboard Brand taxonomy Meta performance memory 500MB uploads
Enterprise
Custom
Custom volume + pricing
Custom onboarding + integrations
Everything in Scale SLA Custom integrations Dedicated account manager
Contact us →
Credit costs
Image = 1 credit Email = 1 credit Carousel = 1 credit Landing Page = 1 credit Video = 2 credits Long Video = 4 + 1/2min over 10min

Drop-in for any AI stack

MCP server is the recommended path. REST API, Help Docs, and launch notes for everything else.

MCP Server →
Claude Desktop, Cursor, ChatGPT, any MCP client.
Help Docs
Set up taxonomy, Meta performance, chat, and MCP.
REST API
Interactive Swagger docs.
Blog
Open taxonomy, AI memory, and creative performance notes.

Give your AI a
creative brain

Taxonomy, memory, and a strategist that runs inside whatever AI you already use. Free to start. Pay per analysis.

Get Your API Key See MCP Setup