Scans
Create a scan
Submit content to Sapient and get a per-second brain-response scan back.
POST
Submit a piece of content (a video, audio track, page URL, or raw text) and Sapient runs it through the brain model, returning a
queued scan you poll (or receive on a webhook_url). The completed scan is a per-second timeline: every lens you ask for, scored second by second, with optional per-second raw brain-network values, benchmarks, reasons, and detected moments.
Pick the lenses you care about and the model to run, either mary (default) or qualia. Completed scans are priced per model: $0.50 for Mary, $1.00 for Qualia (see pricing).
Body Parameters
The content to analyze.
Which brain model to run:
mary (default, $0.50/scan) or qualia (the video specialist, $1.00/scan). Omit the field to use Mary. qualia requires a video URL input. See Models.Which lenses to score. Any subset of the seven:
attention, purchase_intent, manipulation, emotion, cognitive_effort, memory, surprise. Defaults to the top three. Unknown values are ignored.The timeline resolution. Per-second is the default and the only resolution today.
Toggles for what the completed scan includes.
If set, Sapient
POSTs the completed scan to this URL once it finishes. Best-effort; polling is always available as a fallback.Response
The scan identifier, e.g.
mary_run_mq7c8f31_jad6o8tl. Use it to poll for the result.queued on submit. Later transitions to processing, then complete (or error).The normalized lens list this scan will score. It echoes your request, deduped and defaulted.