Url to markdown API
One GET request turns any public web page into clean Markdown with title, description, and published time — no scraping stack to maintain.
- LLM-ready output
- Metadata included
- Idempotent retries
- Per-success pricing
Request examples
curl "https://api.agentbody.io/v1/scraping/jina-reader?url=YOUR_URL" \
-H "Authorization: Bearer <YOUR_AGENTBODY_API_KEY>"Response examples
Documented responses from the OpenAPI specification.
Readable public web page content.
{
"content": "example_value",
"description": "example_value",
"published_time": "example_value",
"title": "example_value",
"url": "https://agentbody.io/"
}Features
Built for feeding web content to LLMs and pipelines.
LLM-ready output
Readable Markdown instead of raw HTML — feeds straight into prompts and RAG.
Metadata included
Title, description, published time, and canonical URL come back alongside content.
Idempotent retries
Retry safely: idempotency keys replay stored responses instead of double-charging.
Per-success pricing
Credits settle per successful request, governed by the live tool price table.
How to use
From zero to clean Markdown in three steps.
Create an API key
Sign up and generate a key in the console — takes under a minute.
Copy a request sample
Grab the curl, JavaScript, Python, Java, or Go sample with your key filled in.
Feed the Markdown
Pipe content into your prompt, RAG index, or CMS — HTML noise is already stripped.
Frequently asked questions
What does the API return?
A JSON object with the page's title, description, published time, canonical URL, and Markdown content.
Does it work with any URL?
Any publicly reachable HTTPS page. Authenticated, private, or redirect-only URLs are rejected.
How is the API priced?
Per successful request, read live from the gateway's tool price table — nothing is hardcoded.
Is the output cleaned?
Yes — boilerplate and navigation noise are stripped, leaving readable Markdown.
Do retries double-charge?
No. Send an idempotency key and retries replay the stored response without a second charge.
Can I parse PDFs with it?
For PDF and document parsing, use the Documents API instead — it has page-range retrieval built in.