APIs

Markdown Generation

Convert HTML to clean, LLM-friendly markdown with intelligent content filtering

Overview

The Markdown Generation API converts web pages to clean markdown perfect for LLM processing and RAG systems. Powered by Crawl4AI, it provides intelligent content filtering.

Quick Start

curl -X POST https://api.ujeebu.com/markdown \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "url": "https://blog.example.com/article"
  }'

import { UjeebuClient } from '@ujeebu-org/ujeebu-sdk';

const client = new UjeebuClient('YOUR_API_KEY');

const response = await client.markdown(
  'https://blog.example.com/article'
);

console.log(response.data.markdown);

from ujeebu_python import UjeebuClient

client = UjeebuClient('YOUR_API_KEY')

response = client.markdown(
    'https://blog.example.com/article'
)

data = response.json()
print(data['markdown'])

client, _ := ujeebu.NewClient("YOUR_API_KEY")

result, _, err := client.Markdown(ujeebu.MarkdownParams{
    URL: "https://blog.example.com/article",
})

if err != nil {
    log.Fatal(err)
}

fmt.Println(result.Markdown)

Parameters

Parameter	Type	Required	Default	Description
`url`	`string`	Yes	`-`	URL to convert to markdown
`filter`	`string`	No		Content filter: 'raw', 'fit' (default), 'bm25'
`query`	`string`	No		Query for BM25 filtering
`citations`	`boolean`	No	`true`	Generate numbered citations. Defaults to `false` on POST requests.
`js`	`boolean`	No	`true`	Execute JavaScript.
`proxy_type`	`string`	No		Proxy type: 'rotating', 'advanced', 'premium', 'residential', 'residential_us', 'residential_geo'. If not set, auto_proxy selects the best proxy automatically.
`wait`	`number`	No		Wait time in ms after page load
`wait_for_selector`	`string`	No		CSS selector to wait for before conversion
`proxy`	`string`	No		Custom proxy URL to use for the request
`timeout`	`number`	No	`60`	Timeout in seconds.
`auto_captcha_solve`	`boolean`	No	`true`	Enable automatic CAPTCHA detection and solving.
`auto_captcha_solve_timeout`	`number`	No		Timeout in milliseconds for CAPTCHA solving

Filter Types

Raw Filter

Complete page markdown without filtering.

curl -X POST https://api.ujeebu.com/markdown \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "url": "https://example.com",
    "filter": "raw"
  }'

import { UjeebuClient } from '@ujeebu-org/ujeebu-sdk';

const client = new UjeebuClient('YOUR_API_KEY');

const response = await client.markdown(
  'https://example.com',
  { filter: 'raw' }
);

console.log(response.data.markdown);

from ujeebu_python import UjeebuClient

client = UjeebuClient('YOUR_API_KEY')

response = client.markdown(
    'https://example.com',
    params={'filter': 'raw'}
)

data = response.json()
print(data['markdown'])

client, _ := ujeebu.NewClient("YOUR_API_KEY")

result, _, err := client.Markdown(ujeebu.MarkdownParams{
    URL:    "https://example.com",
    Filter: "raw",
})

if err != nil {
    log.Fatal(err)
}

fmt.Println(result.Markdown)

Fit Filter (Default)

Extracts main content using Trafilatura/Readability.

curl -X POST https://api.ujeebu.com/markdown \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "url": "https://blog.example.com",
    "filter": "fit"
  }'

import { UjeebuClient } from '@ujeebu-org/ujeebu-sdk';

const client = new UjeebuClient('YOUR_API_KEY');

const response = await client.markdown(
  'https://blog.example.com',
  { filter: 'fit' }
);

console.log(response.data.fit_markdown);

from ujeebu_python import UjeebuClient

client = UjeebuClient('YOUR_API_KEY')

response = client.markdown(
    'https://blog.example.com',
    params={'filter': 'fit'}
)

data = response.json()
print(data['fit_markdown'])

client, _ := ujeebu.NewClient("YOUR_API_KEY")

result, _, err := client.Markdown(ujeebu.MarkdownParams{
    URL:    "https://blog.example.com",
    Filter: "fit",
})

if err != nil {
    log.Fatal(err)
}

fmt.Println(result.FitMarkdown)

BM25 Filter

Relevance-based filtering with query.

curl -X POST https://api.ujeebu.com/markdown \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "url": "https://docs.example.com",
    "filter": "bm25",
    "query": "installation instructions"
  }'

import { UjeebuClient } from '@ujeebu-org/ujeebu-sdk';

const client = new UjeebuClient('YOUR_API_KEY');

const response = await client.markdown(
  'https://docs.example.com',
  { filter: 'bm25', query: 'installation instructions' }
);

console.log(response.data.markdown);

from ujeebu_python import UjeebuClient

client = UjeebuClient('YOUR_API_KEY')

response = client.markdown(
    'https://docs.example.com',
    params={'filter': 'bm25', 'query': 'installation instructions'}
)

data = response.json()
print(data['markdown'])

client, _ := ujeebu.NewClient("YOUR_API_KEY")

result, _, err := client.Markdown(ujeebu.MarkdownParams{
    URL:    "https://docs.example.com",
    Filter: "bm25",
    Query:  "installation instructions",
})

if err != nil {
    log.Fatal(err)
}

fmt.Println(result.Markdown)

Response

{
  "success": true,
  "url": "https://blog.example.com/article",
  "markdown": "# Full content...",
  "fit_markdown": "# Filtered content...",
  "markdown_with_citations": "# Content⟨1⟩...",
  "references": "## References\\n\\n⟨1⟩ https://example.com",
  "metadata": {
    "html_length": 15420,
    "markdown_length": 3200,
    "conversion_time_ms": 150,
    "filter": "fit"
  }
}

Credits & Billing

Credit Cost

Credit cost per request depends on the proxy type and whether JavaScript rendering is enabled:

Proxy Type	JS Enabled (default)	JS Disabled
rotating (default)	5	1
advanced	10	5
premium	12	8
residential	30	20
residential_us	30	20
residential_geo	10 + 10/MB over 1MB	8 + 10/MB over 1MB

INFO — Credits Header

Consumed credits are returned in the Ujb-credits response header.

INFO — Auto Proxy

If no proxy_type is specified, auto proxy is enabled by default. The system automatically tries different proxies until one succeeds. Credits are charged based on the proxy that was actually used for the successful request.

Cost Example

# Convert a page with default proxy (rotating, js=true) — 5 credits
curl -X POST https://api.ujeebu.com/markdown \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"url": "https://example.com/page"}'

# Convert a page with residential proxy — 30 credits
curl -X POST https://api.ujeebu.com/markdown \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"url": "https://example.com/page", "proxy_type": "residential"}'

import { UjeebuClient } from '@ujeebu-org/ujeebu-sdk';

const client = new UjeebuClient('YOUR_API_KEY');

// Convert 100 pages with default proxy (rotating, js=true)
for (const url of pageUrls) {
  await client.markdown(url);
}
// Total: 100 × 5 = 500 credits

// Convert 100 pages with residential proxy
for (const url of pageUrls) {
  await client.markdown(url, { proxy_type: 'residential' });
}
// Total: 100 × 30 = 3,000 credits

from ujeebu_python import UjeebuClient

client = UjeebuClient('YOUR_API_KEY')

# Convert 100 pages with default proxy (rotating, js=true)
for url in page_urls:
    client.markdown(url)
# Total: 100 × 5 = 500 credits

# Convert 100 pages with residential proxy
for url in page_urls:
    client.markdown(url, params={'proxy_type': 'residential'})
# Total: 100 × 30 = 3,000 credits

client, _ := ujeebu.NewClient("YOUR_API_KEY")

// Convert 100 pages with default proxy (rotating, js=true)
for _, url := range pageURLs {
    _, _, err := client.Markdown(ujeebu.MarkdownParams{URL: url})
    if err != nil {
        log.Println(err)
    }
}
// Total: 100 × 5 = 500 credits

// Convert 100 pages with residential proxy
for _, url := range pageURLs {
    _, _, err := client.Markdown(ujeebu.MarkdownParams{
        URL:       url,
        ProxyType: "residential",
    })
    if err != nil {
        log.Println(err)
    }
}
// Total: 100 × 30 = 3,000 credits

Billing Notes

Credits charged only on successful conversion
Failed requests (4xx, 5xx errors) are not charged
residential_geo proxy requests exceeding 1MB incur an additional 10 credits per MB
When using auto_proxy, only the successful proxy attempt is billed
When a CAPTCHA is detected and solved, an additional +5 credits surcharge is applied on top of the base request cost (auto_captcha_solve is enabled by default)

Next Steps

Check AI Scraper for intelligent extraction

Ready to build?

Spin up an API key in 60 seconds

Free tier: 5,000 credits, no card, full access to every endpoint on this page.

Get free API key or try the playground →