[Bug]: llm_strategy not working

crawl4ai version

Version: 0.4.248

Expected Behavior

Expecting result.extracted_content based on my Pydantic model.

Current Behavior

Code is not reaching the OpenAI LLM successfully on Mac OS. I confirmed by using a Python script without crawl4ai code to openai LLM and execute and got a successful result.

Is this reproducible?

Yes

Inputs Causing the Bug

Steps to Reproduce

run code and program returns None in result.extracted_content

Code snippets

from pydantic import BaseModel, Field
from crawl4ai.extraction_strategy import LLMExtractionStrategy
from crawl4ai  import AsyncWebCrawler, CacheMode
from typing import List, Optional
import asyncio
import os
from dotenv import load_dotenv
import json
from crawl4ai.async_configs import BrowserConfig, CrawlerRunConfig

load_dotenv("/Users/stevenheadley/.zshrc")

# Define schema
class Listing(BaseModel):
    Title: str = Field(..., description="The title of the listing")
    Price: str = Field(..., description="The price of the listing")
    Bedrooms: str = Field(..., description="The number of bedrooms for the listing")
    Bathrooms: str = Field(..., description="The number of bathrooms for the listing")
    Area: str = Field(..., description="The square foortage of the listing")
    Description: str = Field(..., description="The description of the listing")
    Features: str = Field(..., description="The features of the listing")
    content: Optional[str] = Field(..., description="The completecontent of the listing.")

# Create strategy
strategy = LLMExtractionStrategy(
    provider="openai/gpt-4o",
    api_token=os.getenv("OPENAI_API_KEY"),
    schema=Listing.model_json_schema(),
    extraction_type='schema',
    instruction=f"Extract 'Title', 'Price', 'Bedrooms', 'Bathrooms', 'Area', 'Description', and 'Feature' from the content and return the list of listing.",
    verbose=True
)

async def main():
    browser_config = BrowserConfig(viewport_width=1280, viewport_height=720,browser_type="chromium", headless=True, verbose=True)  # or False to see the browser
    
    run_config = CrawlerRunConfig(
        cache_mode=CacheMode.BYPASS,
        remove_overlay_elements=True,
        exclude_external_links=True,
        exclude_social_media_links=True,
        excluded_tags=["nav", "header", "footer"],
        remove_forms=True,
        verbose=True,
        session_id='1234'
    )

   

    async with AsyncWebCrawler(config=browser_config) as crawler:
        result = await crawler.arun(
            url="https://barrierreefrealty.com/property/coconuts-caribe-phase-ii-2-bedroom/",
            config=run_config,
            verbose=True,
            extraction_strategy=strategy
        )
#        data = json.loads(result.extracted_content)
        print(result.extracted_content)  # Print clean markdown content

if __name__ == "__main__":
    asyncio.run(main())
    # Access extracted data
#    data = json.loads(result.extracted_content)

OS

MacOS

Python version

3.9.4

Browser

Chrome

Browser version

132.0.6834.83

Error logs & Screenshots (if applicable)

No errors, just empty results

[frontenders-game]

You should use extraction_strategy with run_config

    run_config = CrawlerRunConfig(
        cache_mode=CacheMode.BYPASS,
        remove_overlay_elements=True,
        exclude_external_links=True,
        exclude_social_media_links=True,
        excluded_tags=["nav", "header", "footer"],
        remove_forms=True,
        verbose=True,
        session_id='1234',
        extraction_strategy=strategy

    )

...

    async with AsyncWebCrawler(config=browser_config) as crawler:
        result = await crawler.arun(
            url="https://barrierreefrealty.com/property/coconuts-caribe-phase-ii-2-bedroom/",
            config=run_config,
            verbose=True,
        )

result:

Extracted items: [{'Title': 'Coconuts Caribe Phase II 2 Bedroom', 'Price': '$450,000', 'Bedrooms': '2', 'Bathrooms': '2', 'Area': '1616 sq ft', 'Description': "Phase I was a tremendous success and we are proud to announce Phase II with construction currently in progress. This fifth-floor condo is now available with a financing option at 8% interest over 30 Years with only a $100,000.00 USD downpayment. Contact us to find out more about the floor plan, and layout, and view a similar interior finish. There are only a few condos, so don't miss out on this great opportunity. This condo has amazing views of the sunset and lagoon side of the island. Coconuts Caribe II is within easy access to many of the popular restaurants on the island and is right along the main paved road. The property is located just slightly outside of the main streets of San Pedro Town.", 'Features': 'Back-up Generator, Elevator, Swimming pool', 'content': None, 'error': False}]

[aravindkarnam]

@frontenders-game Thanks for making the right suggestion. @sheadley We've moved all strategies inside the CrawlerRunConfig and the old legacy params will be deprecated soon. So going forward pass the extraction_strategy to CrawlerRunConfig.

Closing this issue.