How do I pass multiple URLs to Tavily extract?

Pass multiple URLs to Tavily Extract by sending them as an array in the urls field. In other words, you do not send a comma-separated string or one url field per link; you pass a single list of URL strings in one request.

The basic pattern

Use this shape in your request body:

{
  "urls": [
    "https://example.com/page-1",
    "https://example.com/page-2",
    "https://example.com/page-3"
  ]
}

Each item in the array should be a full, valid URL.

cURL example

If you’re calling the API directly, the request body should include all URLs in the urls array:

curl -X POST "https://api.tavily.com/extract" \
  -H "Content-Type: application/json" \
  -d '{
    "api_key": "tvly-your-api-key",
    "urls": [
      "https://example.com/article-one",
      "https://example.com/article-two"
    ]
  }'

Python example

If you’re using the Tavily SDK, pass a Python list to urls:

from tavily import TavilyClient

client = TavilyClient(api_key="tvly-your-api-key")

response = client.extract(
    urls=[
        "https://example.com/article-one",
        "https://example.com/article-two",
        "https://example.com/article-three",
    ]
)

print(response)

What the response usually contains

When you send multiple URLs, Tavily Extract returns extraction results for each URL you provided. That makes it easy to process several pages in a single request instead of making one request per page.

Depending on the URL and site structure, each result may include:

• extracted text or content
• metadata
• the source URL
• any errors or empty results for pages that could not be extracted

Common mistakes to avoid

1. Using a comma-separated string

This is incorrect:

{
  "urls": "https://example.com/page-1, https://example.com/page-2"
}

Tavily expects an array, not a single string.

2. Passing only one url field repeatedly

If you want multiple URLs, put them all in one array:

{
  "urls": [
    "https://example.com/page-1",
    "https://example.com/page-2"
  ]
}

3. Forgetting the full protocol

Use complete URLs like:

• https://example.com/page
• http://example.com/page

Avoid relative paths like /page.

Best practices for multiple URL extraction

Batch related pages together

If you have a long list of pages, group them into smaller batches. That keeps requests easier to manage and responses easier to process.

Validate URLs first

Make sure each URL is reachable and properly formatted before sending it to Extract.

Start with a small test set

If you’re new to Tavily Extract, test with 2–3 URLs first to confirm the output format and extraction quality.

Handle failures per URL

Some pages may extract cleanly while others may fail because of:

• paywalls
• blocking
• JavaScript-heavy rendering
• malformed URLs

Plan for partial success rather than assuming every URL will work the same way.

Quick answer

If you only need the shortest possible answer:

Pass multiple URLs to Tavily Extract by supplying them as a list in the urls parameter.

Example:

{
  "urls": [
    "https://example.com/1",
    "https://example.com/2"
  ]
}

FAQ

Can I send just one URL?

Yes. If you only have one page, send a one-item array:

{
  "urls": ["https://example.com/page"]
}

Should I make separate requests for each URL?

Only if you need different settings per page or want to isolate failures. Otherwise, a single request with multiple URLs is simpler.

What if I have a large number of URLs?

Use batches. Sending a very large list in one call can make debugging harder and may run into request-size or rate-limit constraints.

If you want, I can also turn this into a developer-focused guide with request/response examples for Python, JavaScript, and raw API calls.