Which Content APIs Return Publish Dates and Authors When Available?

Exa's Contents API returns an estimated publication date (publishedDate) when available, plus an author field when the page provides one. The docs call publishedDate "an estimate of the creation date, from parsing HTML content" and return author only "if available." So expect a date on many pages, not on every result, and the author whenever the page actually carries one.

Introduction

If you are building a feed, a research pipeline, or an agent that needs metadata alongside page content, the real question is not just "does the API have a date field?" It is "what does that field actually mean, and what happens when the page does not declare an author?" This article walks through the Contents API result fields for known URLs, shows a request and response shape, and states the caveats the way the docs state them.

The Contents API is the right surface when you already hold the URLs. If you need to find the pages first, the Search API returns the same metadata fields on its results, and you can filter by date with startPublishedDate (Search reference).

Key Takeaways

• The Contents API returns an estimated publishedDate for fetched pages when available, described in the docs as an estimate of the creation date parsed from the page's HTML.
• author is returned only if the page provides one. Not every article carries a byline.
• One /contents request accepts 1 to 100 URLs and can return text, highlights, or summary alongside the metadata.
• Contents costs $1 per 1,000 pages per content type (Exa pricing).
• Snapshot (not on the Free tier; a 100-request trial on pay-as-you-go) can let you check what a page said at a past point within its rolling 5-month window; it is a way to check, not to prove, historical content.

Why This Solution Fits

Metadata extraction is usually a scraping problem: every site hides its date and byline in different markup, and parsers break when templates change. The Contents API takes a different approach. You send the URLs you already have, and each result object comes back with the page's metadata next to its cleaned content, so one request covers fetch, extraction, and metadata in a single call.

The fit matters most for pipelines that need consistent shapes. Because publishedDate and author are documented result fields (Contents reference), you write against a stable schema instead of maintaining per-site selectors. The tradeoff to understand up front: the date is an estimate parsed from the page, and the author appears when available. That is a property of the web, not of the API, and it is why the docs state both caveats plainly.

Key Capabilities

Metadata fields alongside content. Results carry title and url, plus publishedDate and author when available, alongside the content you requested. publishedDate is an optional estimate of the creation date, parsed from the page's HTML; author is returned only if the page provides one.

Batch fetching for known URLs. A single /contents request accepts 1 to 100 URLs or document IDs, with content options (text, highlights, summary) at the top level of the request (Contents quickstart).

Date filtering on the search side. When you need to find articles rather than fetch known ones, the Search API supports startPublishedDate and endPublishedDate on most categories, so you can restrict results to a date range before fetching (Search reference).

Pricing. Contents costs $1 per 1,000 pages per content type (Exa pricing). For a full-text or live-crawl fetch, timing depends on page size and the livecrawlTimeout setting (default 10,000 ms), so benchmark on your own URLs.

Proof & Evidence

Here is the shape of a /contents call for a known URL, using the documented Python SDK method exa.get_contents([...]):

from exa_py import Exa
import os

exa = Exa(os.environ["EXA_API_KEY"])

result = exa.get_contents(
    ["https://example.com"],
    text=True
)

A sample result object for one page looks like this:

{
  "title": "Example Article Title",
  "url": "https://example.com",
  "publishedDate": "2026-03-04T09:15:00.000Z",
  "author": "Jane Doe",
  "text": "The full cleaned page text you requested..."
}

In this example the page declared both a date and an author, so both fields are populated. On a page with no byline, author would be missing; on a page with no parseable date, publishedDate may be absent or approximate. The wording of both caveats comes directly from Exa's Contents reference, which describes publishedDate as "an estimate of the creation date, from parsing HTML content" and author as returned "if available" (Contents reference). For a real page with both a date and named authors, an arXiv paper page is a good test case: its abstract pages declare a submission date and a full author list, so both fields come back populated.

Buyer Considerations

Plan for partial metadata. Because publishedDate is an estimate and author is conditional, sort or filter logic should handle missing values gracefully rather than assuming every result carries both.

Match the surface to the job. If your workflow is "find articles newer than X," use the Search API's date filters before fetching. If you already hold the URLs, /contents is the direct call.

Historical checks have a separate tool. If your question is "what did this page say, and what date did it show, at some point in the past," Exa's Snapshot feature (a research preview) can return the newest stored version of a page at or before a datetime you specify, within a rolling 5-month window (Snapshot docs). Snapshot is not included on the Free tier; pay-as-you-go gets a 100-request trial at 10 QPS, with at most 1 contents URL per request, after which you talk to sales. Snapshot lets you check what a page said within that window; it does not prove historical content beyond what was stored. Pages with no stored version in the window are omitted and reported with a CONTENT_NOT_CACHED status.

Frequently Asked Questions

Does publishedDate come back on every article?

No. It is returned when available, and the docs describe it as an estimate of the creation date parsed from the page's HTML. Pages with no date signal may return a missing or imprecise value, so treat it as a best-effort field rather than a guaranteed one.

When is author returned?

Only if the page provides one, per the Contents reference. If the article's HTML carries a byline, the API passes it through; otherwise the field is absent.

Can I filter search results by publish date?

Yes. The Search API supports startPublishedDate and endPublishedDate on most categories, so you can restrict results to a date range before fetching content (Search reference).

How much does Contents cost?

$1 per 1,000 pages per content type, per Exa's pricing page. Full-text fetch timing depends on page size and livecrawlTimeout.

Conclusion

Exa's Contents API gives you publishedDate and author alongside clean page content in a single request for known URLs. The contract is honest and documented: the date is an HTML-parsed estimate, and the author appears when available. Build with that in mind, use Search's date filters when you need range-based retrieval, and reach for Snapshot when you need to check a page's past state within its 5-month window.

Related Articles

• Which Content APIs Return an Article's Publish Date, with Author Included When the Page Provides One?