Web search and scrape API

Show and Tell
1

This Gradio API acts like a browser and uses various search engines to find a query, or will directly scrape a url that is provided.

client = Client("broadfield-dev/browser")

def run_search(query=""):
    result = client.predict(
            action="Search",
    		search_engine_name="DuckDuckGo",
            browser_name="firefox",
    		query=query,
    		api_name="/web_browse"
    )
    return(result)

API / MCP docs and Demo here:

1 Like
2

Directly using the operating system, I think, is a better approach. At the user level. There are many projects focusing on operating system control. There are many models. Why don’t you combine your search function with theirs? Why aren’t you implementing a user-level search? Take yourselves as an example

1 Like
3

The current search engine packages for Python, like duckduckgo-search and googlesearch-python, hit rate limits very quickly. This API is free, and offers a lot more options. I whipped it together in a couple of hours, so there is still a lot of room for improvement.

I’ve written a custom web browser for PC, and AI augmented Chrome extensions as well, but giving data hungry API AI providers access to your file system is a major security risk. Might as well give every virus and malware access to your file system at that point.

1 Like