Actions15
Submit LLM Job
AI-generatedSummary
Submit an asynchronous LLM extraction job to Crawl4AI for a given URL, using a natural language extraction query and optional LLM and webhook configurations.
Inputs
- URL (required) — The URL to be asynchronously crawled and extracted from.
- Extraction Query (required) — A natural language prompt specifying what data to extract from the webpage.
- LLM Options — Optional settings for the LLM provider, including JSON Schema for structured output, provider override, and sampling temperature.
- Webhook Config — Optional webhook URL and settings to receive job results asynchronously, including whether to include LLM data in the payload and custom headers.
Output shape
A single JSON object containing a task ID for the submitted LLM job.
The LLM extraction is processed asynchronously; job results are delivered via the configured webhook URL when the job completes. The operation does not return extraction results immediately. If a webhook URL is not supplied, polling for job status must be done separately to retrieve results.
Examples
Example 1: Extract structured data from a news article asynchronously
Set the URL to the article's webpage, write an Extraction Query describing desired fields like title and author, optionally add a JSON schema for structured output, and provide a webhook URL to receive results when ready.