# Multi-URL Extract

> Structured extraction across many URLs with optional SearXNG grounding — 4 credits per URL.

Source: http://localhost:3000/docs/endpoints/multi-extract

## Multi-URL Extract: `POST /v1/multi-extract/`

One schema applied across many URLs, optionally grounded by search. Cost: 4 credits.

```bash
curl -X POST https://api.hushsearch.io/v1/multi-extract/ \
  -H "X-API-Key: hs_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "urls": [
    "https://fastapi.tiangolo.com/",
    "https://www.djangoproject.com/"
  ],
  "prompt": "Project name and one-line description"
}'
```

**POST** `/v1/multi-extract/` — **4 credits per URL** (max 50 URLs)

Extract structured data from multiple URLs in parallel. Optionally ground URL discovery with a SearXNG query when `urls` is empty or sparse.

## Request

```json
{
  "urls": [
    "https://example.com",
    "https://openai.com"
  ],
  "schema_fields": ["title", "description"],
  "prompt": "Extract page metadata",
  "ground_query": "OpenAI company homepage",
  "num_results": 5
}
```

| Field | Type | Required | Description |
|---|---|---|---|
| `urls` | list[string] | No* | Up to 50 URLs to extract |
| `ground_query` | string | No* | SearXNG query to discover URLs |
| `schema_fields` | list or JSON schema | No | Fields to extract per URL |
| `prompt` | string | No | Custom extraction instruction |
| `num_results` | int | No | Results when grounding via search |

\* Provide `urls` and/or `ground_query`.

## Response

```json
{
  "results": [
    {
      "url": "https://example.com",
      "data": { "title": "Example Domain", "description": "..." },
      "success": true
    }
  ],
  "success_count": 2,
  "url_count": 2,
  "grounded": false,
  "credits_used": 8,
  "credits_remaining": 92,
  "response_time_ms": 3100
}
```

## Notes

- Credit cost scales with URL count: `4 × min(url_count, 50)`.
- For long-running batches, use [Async Jobs](http://localhost:3000/docs/endpoints/jobs) with `type: "multi_extract"`.
