# Find Similar

> Find pages similar to a URL or a passage of text.

Source: http://localhost:3000/docs/endpoints/similar

## Find Similar: `POST /v1/search/similar`

Pages semantically similar to a URL or a passage of text. Cost: 1 credit. Available on the free demo.

```bash
curl -X POST https://api.hushsearch.io/v1/search/similar \
  -H "X-API-Key: hs_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "url": "https://fastapi.tiangolo.com/",
  "num_results": 5
}'
```

Give a URL or a block of text and get back pages that cover the same ground, ranked by semantic similarity.

## Parameters

| Field | Type | Default | Notes |
|---|---|---|---|
| `url` | string | — | Page to find neighbours for. Provide `url` or `text`. |
| `text` | string | — | Passage to find neighbours for. |
| `num_results` | integer | `10` | Number of results. |
| `mode` | string | — | `neural`, `hybrid` or `personal_first` (prefers your private index). |
| `include_domains` / `exclude_domains` | string[] | — | Restrict or exclude domains. |
| `exclude_source_domain` | boolean | — | Drop results from the source URL's own domain. |
| `start_published_date` / `end_published_date` | datetime | — | Publication date window (ISO 8601). |
| `category` | string | — | Soft category hint. |

## Response

The same shape as [Web Search](http://localhost:3000/docs/endpoints/search): `results[]` with `title`, `url` and `snippet`, plus `credits_used`, `credits_remaining` and `response_time_ms`.
