01 / Schema driven
Ask for exact fields.
Describe the output your workflow expects and receive a predictable JSON object in response.
Product / Extract
Turn web pages into structured JSON.
Define the fields your workflow needs and extract predictable data from unstructured pages.
curl -X POST https://api.beecrawl.dev/extract \
-H "X-Web-Extract-Api-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"url":"https://example.com",
"schema":{
"company":"Company name",
"email":"Contact email"
}
}'Features // Product / Extract
Move from page content to application-ready fields without maintaining a custom parser for every source.
02 / Flexible extraction
Choose how data is parsed.
Use deterministic page parsing by default or connect an OpenAI-compatible model for complex extraction.
03 / Browser ready
Handle dynamic sources.
Use browser rendering when the values you need only appear after the page has loaded.
How it works //
From request to ready data.
01
Send a URL and field schema.
02
Render, read, and match the page.
03
Receive structured data for your workflow.
{
"requestId": "req_01H...",
"status": "ready",
"data": "useful web content",
"metadata": { "provider": "beecrawl" }
}FAQ //
Everything you need to get moving.
How do I define the output?
Pass a schema whose keys name the fields you need and whose values describe the data BeeCrawl should extract.
Does Extract require an LLM?
No. BeeCrawl uses deterministic page parsing by default and can use a configured OpenAI-compatible provider for more complex extraction.