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

# Effort levels

> Choose how much work goes into ranking, and what a search costs.

Every search scores each article on relevance to your request, the authority
of its publisher, how many newsrooms cover its story and how fresh it is.
`effort` decides how much further work goes into the order.

| Effort | What it adds | Credits |
| - | - | - |
| `low` | Nothing: the first ranking is the answer. Fastest. | 1 NCC |
| `medium` (default) | A cross-encoder rerank of the top stories. | 5 NCC |
| `high` | An LLM judgement of the top stories, and merges stories that report the same event. Slowest, best quality. | 10 NCC |

```bash theme={null}
curl -X POST "https://api.newscatcherapi.com/news-api/search" \
  -H "x-api-key: $NEWSCATCHER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"q": "EU AI Act enforcement", "effort": "high"}'
```

The response repeats the effort that ran in `effort`, and how long the search
took in `took_ms`.

## Which effort to use

> [accordion] low: high volume, keyword-style queries
>
> Feeds, alerts and dashboards that send many searches with clear keywords,
> where speed and price matter more than the exact order.

> [accordion] medium: the default for most requests
>
> Questions in plain words, user-facing search and agent tool calls.

> [accordion] high: when the top results must be right
>
> Research, briefings and reports, where a wrong top result costs more than
> the extra seconds. Stories about the same event are merged, so the top
> results cover more distinct events.

## Rate limits per effort

Your plan sets requests per second for each effort level. See
[Rate limits](/docs/get-started/rate-limits.md).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.