# Staan Fetch: a URL to Markdown API for LLMs and AI agents

_2026-11-17 · Product updates · By Loïc Journée_

![Staan Fetch reads the URL eu-searchperspective.com and returns its content as Markdown](https://staan.ai/assets/cover-CspppCV0.jpg)

Today, we're launching Fetch, the third API in the Staan family. Staan Fetch is a URL to Markdown API: send up to 10 URLs in one call, and get back the main content of each page as clean Markdown or plain text, ready for an LLM. It reads web pages, PDF and DOCX files, from Staan's European index or live. Fetch is available today on request.

## What is a URL to Markdown API?

A URL to Markdown API takes the address of a web page and returns its content as Markdown, without the code around it. For an LLM, that's the useful part of the page: headings, paragraphs, lists and tables, without navigation, scripts or styling.

Markdown keeps the structure of a page in a compact, readable form. That's why it has become a common input format for RAG pipelines and AI agents.

## Why we built Fetch

Our search APIs help AI agents find the right pages. But a list of links is rarely the answer: the answer is inside the pages. To get to it, teams have had to run a scraper next to their search API, and keep it working: downloading pages, stripping layouts, opening PDFs, retrying slow sites. Fetch replaces that plumbing with one call.

## How do you convert a URL to Markdown with Staan Fetch?

Send the URLs you want to read to the `/v1/fetch` endpoint, with your API key:

```bash
curl -X POST "https://api.staan.ai/v1/fetch" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"urls": ["https://eu-searchperspective.com/"]}'
```

Each page comes back with its content, its title, where it came from and when it was retrieved (response abridged):

```json
{
  "results": [
    {
      "url": "https://eu-searchperspective.com/",
      "title": "European Search Perspective",
      "content": "We believe in plurality and diversity of thought.\n\nWe believe in the power of competition and choice.\n\n…",
      "content_format": "markdown",
      "file_type": "html",
      "source": "live",
      "fetched_at": "2026-11-19T08:00:00.000Z"
    }
  ],
  "errors": [],
  "usage": { "urls_billed": 1 }
}
```

URLs that couldn't be read are listed in `errors`, each with a reason. They don't fail the call, and they aren't billed.

## Cached or live: how fresh is the content?

Fetch reads first from the Staan index, where pages have already been crawled. You decide how recent a copy you accept:

- By default, Fetch serves the indexed copy, and fetches the page live when there is none.
- With `max_age`, you set the oldest version you'll accept, in seconds. Anything older is fetched live.
- With `max_age` set to 0, every page is fetched live.

PDF and DOCX files are always fetched live. Each result says whether it came from the index or from a live fetch, and when.

## What does Fetch cost?

Fetch costs €0.50 per 1,000 URLs, excluding VAT. You're billed for the URLs returned with results:

- failed URLs are never billed;
- a URL sent twice in the same call is billed once;
- a page served from the index costs the same as a page fetched live.

## Which Staan API do you need?

| You start from                       | Use                                                        | You get                               |
| ------------------------------------ | ---------------------------------------------------------- | ------------------------------------- |
| URLs you already have                | [Fetch](https://docs.staan.ai/docs/fetch)                  | The main content of each page         |
| A query, to get links                | [Web Search](https://docs.staan.ai/docs/web-search)        | 10 ranked results as JSON             |
| A query, to get context for your LLM | [Web Search for AI](https://docs.staan.ai/docs/web-for-ai) | Scored snippets and full page content |

## Built on an independent European index

Fetch is hosted in the EU and reads first from Staan's independent European web index, built by European Search Perspective, the joint venture of Qwant and Ecosia. The same index powers Staan's search APIs.

## Learn more

- Every parameter and response field: read the [Fetch documentation](https://docs.staan.ai/docs/fetch).
- Formats, freshness and billing: see the [FAQ](/faq).
- New to Staan? [Get your free API key](/auth/register) and start with Web Search.

---

**Get access to Fetch** Fetch is available on request. Tell us about your use case and volumes, and we will enable it on your account.

[Talk to sales](https://form.typeform.com/to/xbbrneRe) · [Read the docs](https://docs.staan.ai)
