Text Extraction

​

Basic Usage

from askui import VisionAgent

with VisionAgent() as agent:
    agent.tools.webbrowser.open_new("http://www.example.com")
    agent.wait(3)
    
    text = agent.get("What is the main heading?", response_schema=str)
    print(f"main heading: {text}")

​

Best Practices

1. Be Specific About Location: Mention where the text is located
   # Good - specific location
   header = agent.get("What is the text in the page header?", response_schema=str)
   
   # Less specific
   text = agent.get("What text is shown?", response_schema=str)
   
2. Handle Empty or Missing Text: Consider that text might not exist
   from typing import Optional
   
   # Text might not be present
   optional_text = agent.get("What is the subtitle, if any?", response_schema=Optional[str])
   
   if optional_text:
       print(f"Subtitle: {optional_text}")
   
3. Clean and Validate Extracted Text: Post-process extracted text as needed
   # Extract and clean price
   price_text = agent.get("What is the price?", response_schema=str)
   price_value = float(price_text.replace("$", "").replace(",", ""))