Playground
Compose Crawl4AI requests
A hands-on workspace for building the API and MCP requests that drive crawling, extraction, and artifact generation.
port 11235health / ready / metrics / playgroundweekly buildsamd64 + arm64
/crawl
Create a crawl job
POSTexample
{
"url": "https://docs.crawl4ai.com/",
"strategy": "browser",
"wait_for": "css:main",
"js_code": ["window.scrollTo(0, document.body.scrollHeight)"]
}/crawl/stream
Stream crawl progress
POSTexample
{
"urls": [
"https://docs.crawl4ai.com/core/quickstart/",
"https://docs.crawl4ai.com/core/markdown-generation/"
],
"dispatch": "adaptive",
"include_markdown": true
}/crawl/job
Queue a crawl job with webhook delivery
POSTexample
{
"url": "https://docs.crawl4ai.com/blog/releases/0.7.6/",
"webhook_url": "https://example.com/webhook",
"webhook_mode": "full"
}/deep
Traverse a site with deep crawl settings
POSTexample
{
"urls": [
"https://docs.crawl4ai.com/core/quickstart/",
"https://docs.crawl4ai.com/core/deep-crawling/"
],
"dispatch": "adaptive"
}/llm/job
Queue an LLM extraction job with webhook delivery
POSTexample
{
"url": "https://docs.crawl4ai.com/core/quickstart/",
"prompt": "Summarize the main features",
"webhook_url": "https://example.com/webhook",
"webhook_mode": "notify"
}/session
Persist crawl session state
POSTexample
{
"id": "session_docs",
"urls": ["https://docs.crawl4ai.com/core/session-management/"],
"last_action": "scroll"
}/md/{url}
Generate markdown for a page
GETexample
GET /api/md/https://docs.crawl4ai.com/core/link-media/
/llm/{url}
Run LLM extraction against a URL
GETexample
GET /api/llm/https://docs.crawl4ai.com/advanced/advanced-features/
/result/{url}
Inspect the CrawlResult payload
GETexample
GET /api/result/https://docs.crawl4ai.com/core/link-media/
/links/{url}
Inspect extracted links
GETexample
GET /api/links/https://docs.crawl4ai.com/core/link-media/
/media/{url}
Inspect extracted media
GETexample
GET /api/media/https://docs.crawl4ai.com/core/link-media/
/html/{url}
Inspect cleaned HTML
GETexample
GET /api/html/https://docs.crawl4ai.com/core/quickstart/
/screenshot/{url}
Generate a screenshot artifact
GETexample
GET /api/screenshot/https://docs.crawl4ai.com/core/link-media/
/pdf/{url}
Generate a PDF artifact
GETexample
GET /api/pdf/https://docs.crawl4ai.com/advanced/advanced-features/
/execute_js
Run page JavaScript
POSTexample
{
"url": "https://docs.crawl4ai.com/core/quickstart/",
"script": "window.scrollTo(0, document.body.scrollHeight)"
}/ask
Query the library context
POSTexample
{
"prompt": "What Crawl4AI features are available for agents?"
}/config
Inspect browser, crawler, and LLM defaults
GETexample
GET /api/config
/mcp/sse
Inspect the MCP event stream
GETexample
GET /api/mcp/sse
URL rules
Pattern-based crawl routing
POSTexample
{
"urls": [
"https://docs.crawl4ai.com/core/quickstart/",
"https://docs.crawl4ai.com/blog/releases/0.7.6/",
"https://docs.crawl4ai.com/core/deep-crawling/"
],
"route_by_pattern": true
}Response
Live output from the local mock API
Run a mock request to inspect the response.
Feature map
Runtime controls and outputs
Crawl controls
browser | http | deep strategy
wait_for selectors or JS predicates
custom headers and proxy settings
session reuse and local storage
Extraction
CSS, XPath, and LLM-based schemas
clean markdown vs filtered markdown
links and media extraction
structured JSON output for repeated use
Outputs
HTML, markdown, screenshots, and PDFs
streaming NDJSON for long jobs
webhook delivery for async jobs
MCP tools for external agents