Turn any web page into LLM-ready markdown

Send any URL to Fetch with response_type=markdown and get back the page as clean markdown: headings, lists, links and tables kept, scripts and styling dropped. It’s the fastest way to give an LLM, a RAG pipeline or an agent the content of a page. Uses: Fetch with response_type and Adaptive Stealth Mode (mode=auto)

Ask your AI agent

With the Zenrows MCP server connected to Claude, Cursor or any MCP client, paste this prompt. Your agent picks the right Zenrows tools and settings on its own.

Use Zenrows to read https://www.scrapingcourse.com/ecommerce/product/abominable-hoodie/
as markdown, then summarize the product in three bullet points: name, price, and what it's made for.

Fetch a page as markdown

Set response_type=markdown. Pair it with mode=auto so Zenrows renders JavaScript and switches to premium proxies only when the page needs them, and bills only the configuration that succeeds.

import requests

params = {
    "apikey": "YOUR_ZENROWS_API_KEY",
    "url": "https://www.scrapingcourse.com/ecommerce/product/abominable-hoodie/",
    "mode": "auto",
    "response_type": "markdown",
}

response = requests.get("https://api.zenrows.com/v1/", params=params, timeout=180)
response.raise_for_status()

markdown = response.text
print(markdown)

const params = new URLSearchParams({
  apikey: "YOUR_ZENROWS_API_KEY",
  url: "https://www.scrapingcourse.com/ecommerce/product/abominable-hoodie/",
  mode: "auto",
  response_type: "markdown",
});

const response = await fetch(`https://api.zenrows.com/v1/?${params}`, {
  signal: AbortSignal.timeout(180_000),
});
if (!response.ok) throw new Error(`${response.status}: ${await response.text()}`);

const markdown = await response.text();
console.log(markdown);

curl "https://api.zenrows.com/v1/?apikey=YOUR_ZENROWS_API_KEY&url=https%3A%2F%2Fwww.scrapingcourse.com%2Fecommerce%2Fproduct%2Fabominable-hoodie%2F&mode=auto&response_type=markdown"

zenrows fetch https://www.scrapingcourse.com/ecommerce/product/abominable-hoodie/ --output markdown

The CLI uses mode=auto by default, so it needs no extra flag. Install it with npm install -g @zenrows/cli, see the CLI guide.

Example output

The product section of the response, trimmed:

# Abominable Hoodie

$69.00

This is a variable product called a Abominable Hoodie

## Description

It took CoolTech™ weather apparel know-how and lots of wind-resistant fabric to get the Abominable Hoodie just right.

• Blue heather hoodie.
• Relaxed fit.
• Moisture-wicking.
• Machine wash/dry.

Get plain text instead

When you need words only, for example to count tokens, classify text or feed a model that ignores formatting, set response_type=plaintext. Everything else stays the same.

Python (Requests)

import requests

params = {
    "apikey": "YOUR_ZENROWS_API_KEY",
    "url": "https://www.scrapingcourse.com/ecommerce/product/abominable-hoodie/",
    "mode": "auto",
    "response_type": "plaintext",
}

response = requests.get("https://api.zenrows.com/v1/", params=params, timeout=180)
response.raise_for_status()
print(response.text)

Send the page to an LLM

Markdown drops straight into a prompt. This example keeps the content between clear delimiters so the model knows where the page starts and ends:

Python (Requests)

import requests

params = {
    "apikey": "YOUR_ZENROWS_API_KEY",
    "url": "https://www.scrapingcourse.com/ecommerce/product/abominable-hoodie/",
    "mode": "auto",
    "response_type": "markdown",
}
markdown = requests.get("https://api.zenrows.com/v1/", params=params, timeout=180).text

prompt = f"""Answer using only the page below.

<page>
{markdown}
</page>

Question: What sizes and colors is this product available in?"""

# Send `prompt` to the model of your choice, for example with the Anthropic or OpenAI SDK.
print(prompt[:500])

Customize it

Change	How
Plain text instead of markdown	`response_type=plaintext`, see Plain text
A PDF of the page	`response_type=pdf`, see PDF
Content from a specific country	Add `proxy_country=us` (any ISO code), see Proxy country
Wait for content that loads late	Use `js_render=true` with `wait_for=<selector>` instead of `mode=auto`, see Wait for selector
Structured fields instead of a page	Use `extract=auto`, see Get structured JSON from any page
Hundreds or thousands of pages	Send the URLs as one job, see Scrape a list of URLs

Troubleshooting

The markdown is missing content you see in your browser: the content loads after the first render. Switch to js_render=true and add wait_for with a selector of the missing element.
The markdown starts with navigation links: that’s the page’s own menu. Start reading from the first # heading, or use Extract to get only the fields you need.
The request takes long to return: set your HTTP client timeout to 180 seconds or more, so mode=auto has time to escalate when a page needs it.

Build a RAG knowledge base from a website

Crawl a whole site and turn every page into markdown chunks.

Get structured JSON from any page

Skip parsing: get product, listing or article fields as JSON.

Scrape pages protected by anti-bot systems

Let Adaptive Stealth Mode pick the right configuration per page.

Connect your AI agent to Zenrows

Give Claude, Cursor or your own agent live web access.

Start Here

Foundations Course

Fetch Recipes

Extract Recipes

Crawl Recipes

Batch Recipes

Browser Recipes

Pipelines

Use Cases

By Surface

Production

Turn any web page into LLM-ready markdown

Ask your AI agent

Fetch a page as markdown

Example output

Get plain text instead

Send the page to an LLM

Customize it

Troubleshooting

Build a RAG knowledge base from a website

Get structured JSON from any page

Scrape pages protected by anti-bot systems

Connect your AI agent to Zenrows

​Ask your AI agent

​Fetch a page as markdown

​Example output

​Get plain text instead

​Send the page to an LLM

​Customize it

​Troubleshooting

​Related recipes

Build a RAG knowledge base from a website

Get structured JSON from any page

Scrape pages protected by anti-bot systems

Connect your AI agent to Zenrows

Ask your AI agent

Fetch a page as markdown

Example output

Get plain text instead

Send the page to an LLM

Customize it

Troubleshooting

Related recipes