Convert any web page URL into clean Markdown text, removing noise such as ads and navigation bars. Free, no API Key required.
Use Cases
- The user says "help me read the content of this link" or "summarize this article"
- Need to obtain the full content of a web page for analysis, translation, or summarization
- Need to scrape the main body of a web page during an in-depth research process
- The user shared a URL and wants to understand the content
Execution Method
Must be executed through the nodes tool; using api_calling is prohibited!
Read Web Page Content
Append the target URL to https://r.jina.ai/ :
{
"action": "run",
"command": ["curl", "-sS", "-L", "--max-time", "30", "-H", "Accept: text/markdown", "-H", "X-Retain-Images: none", "https://r.jina.ai/https://example.com/article"],
"output_handling": "full"
}
Search Function
{
"action": "run",
"command": ["curl", "-sS", "-L", "--max-time", "30", "-H", "Accept: application/json", "-H", "X-Retain-Images: none", "https://s.jina.ai/搜索关键词"],
"output_handling": "full"
}
Usage Recommendations
- Prefer for long articles: For short pages, just search for a summary; for long articles, use Jina Reader to get the full content
- Use together with search: First use the search function (
https://s.jina.ai/关键词) to find the URL, then use Jina Reader to extract the main body - Timeout handling：
--max-time 30to control the timeout - Content truncation: Extra-long articles are controlled by output_handling to avoid context overflow
Output Specifications
- The extracted content should preserve the original structure (headings, paragraphs, lists)
- Mark the source URL
- If extraction fails, inform the user and explain the reason
Notes
- Do not use api_calling; Jina Reader is not a registered API
- Do not open Safari/browserto do manual searching; directly use nodes to execute the curl command
- Do not use python3 -c "import httpx; ..."; directly use curl (zero dependencies, preinstalled on all systems)