> ## Documentation Index
> Fetch the complete documentation index at: https://api.thesapientcompany.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Models

> Qualia is the model that runs your scan. Mary is temporarily unavailable.

<Warning>
  Mary is temporarily unavailable — all scans run on Qualia, the default model.
</Warning>

Every scan runs through one of Sapient's brain-response models. Today that model is **Qualia**. You don't need to pick one — omit the [`model`](/api-reference/create-scan#param-model) field on [create](/api-reference/create-scan) and your scan runs on Qualia. It returns the four [lenses](/api-reference/lenses), scored second by second.

| Model              | Reads                           | Sharpest at                             | Price per scan | Status                      |
| ------------------ | ------------------------------- | --------------------------------------- | -------------- | --------------------------- |
| `qualia` (default) | Video and audio (no text input) | Telling visually distinct videos apart  | \$2.50         | **Available**               |
| `mary`             | Video, audio, voice, and text   | The spoken, written, or narrative layer | —              | **Temporarily unavailable** |

Qualia is the default. Omit the `model` field entirely and your scan runs on Qualia.

## Qualia

Qualia is a brain encoder. It takes your content, predicts the response of real human brain networks, and reads that prediction out as the four [lenses](/api-reference/lenses), scored second by second. It reads video (visual and audio together) and audio or voice. It does **not** take text input.

Submit a 30-second reel or ad, and you get back a per-second timeline across all four lenses: where attention holds, where buy-intent builds, where it slips. Qualia is sharpest at separating visually distinct pieces of video, which makes it the one to reach for when you're comparing video creatives and need them clearly told apart.

Use Qualia to:

* Score a reel, TikTok, or video ad before you publish or put spend behind it.
* A/B test video creatives where the variants are visual: different cuts, footage, or pacing.
* Pick the strongest of several video edits before launch.

Submit a video URL. For a comparison, submit two video URLs as one scan each.

## Mary (temporarily unavailable)

Mary is temporarily unavailable — all scans run on Qualia, the default model.

<RequestExample>
  ```bash Qualia (default, model omitted) theme={null}
  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" } }'
  ```

  ```bash Qualia (explicit) theme={null}
  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" },
      "model": "qualia"
    }'
  ```
</RequestExample>

<Note>
  Qualia scans cost \$2.50 each, charged once when the scan completes. You're only charged for a scan that completes. Mary is temporarily unavailable. See [pricing](/api-reference/pricing).
</Note>

<Card title="Use cases and applications" href="/api-reference/use-cases">
  Real ways teams use Qualia: A/B testing video creatives, hook analysis, pre-flight QA, and more.
</Card>