Skip to main content
Every scan is read through one or more lenses, each a 0–100 score (50 is neutral) derived from the seven brain networks. You pick which lenses to score in the lenses array, and each one is returned per second in the timeline. The same lens output shape applies on either model.

​
The seven lenses

Lens (API key)MeasuresKnowledge Base
attentionWhether the viewer is locked in, and where they slip away.Attention
purchase_intentThe approach-vs-avoid pull toward buying.Purchase intent
manipulationHow much it leans on emotional persuasion vs plain reasoning.Manipulation
emotionEmotional valence: the felt pull, positive or negative.Emotion
cognitive_effortHow hard the viewer is working to follow it.Cognitive effort
memoryHow likely it is to stick.Memory
surpriseWhether it violates expectation in a way that registers.Surprise

​
Requesting a subset

Pass any subset in the lenses array. Order is preserved, duplicates are dropped, and unknown values are ignored. When omitted, the default is the top three: attention, purchase_intent, manipulation. 
curl https://www.thesapientcompany.com/api/v1/scans \
  -H "Authorization: Bearer sk_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "input": { "type": "video", "video_url": "https://yourcdn.com/ad.mp4" },
    "lenses": ["attention", "manipulation"]
  }'

{
  "scan_id": "mary_run_mq7c8f31_jad6o8tl",
  "status": "queued",
  "lenses": ["attention", "manipulation"]
}
Each second of the completed timeline then carries only those lenses: 
{
  "second": 0,
  "scores": { "attention": 65, "manipulation": 48 }
}