From URL List to Diligence Notes: Using the Exa Contents API for Private-Company Research

The Exa Contents API turns a list of private-company URLs into readable due-diligence notes. Send up to 100 URLs in one request and get back clean page text, query-targeted highlights, and schema-guided summaries, with cached content served instantly and live crawling as a fallback. At $1 per 1,000 URLs per requested content type, it replaces manual reading.

Introduction

Private-company due diligence is a URL problem before it is an analysis problem. By the time you sit down to write notes, you usually have a folder of links: the company's About and Careers pages, press releases, funding announcements, regulatory filings, podcast transcripts, news coverage, an investor memo PDF. Reading each page in a browser tab and typing up findings is slow, and writing your own scraper for every site is worse, because every site renders differently and breaks differently.

Exa's Contents API is built for exactly this step. Exa's docs say to use /contents when you already know the URLs: you pass the list in, and the API returns full page text cleaned of navigation, ads, and boilerplate, plus optional summaries and highlights, in formats ready to feed into an LLM or paste straight into your memo (Contents API quickstart). This article explains what the tool does, why it fits due-diligence work specifically, and what to check before you build on it.

Key Takeaways

The Exa Contents API accepts 1 to 100 URLs or document IDs per request and returns full text, highlights, summaries, and metadata for each page, using cached content when available and live crawling for uncached pages (Contents reference).
It handles the messy sources diligence actually produces: JavaScript-rendered pages, PDFs such as filings and investor decks, and complex layouts, returned as clean Markdown.
Query-targeted highlights return only the passage that answers your question, with the source URL attached for citation, and highlight extraction takes under 100 ms (Exa vs Parallel).
On /contents, each content type (text, highlights, summary) is billed separately at $1 per 1,000 pages; Exa's free tier gives $10 in credits every month (up to 2,500 Lite searches), plus a one-time $10 onboarding bonus, with no payment method required (Pricing).
Per-URL status reporting means a dead link or failed crawl is flagged explicitly, so your notes can state exactly which sources were reviewed.

Why This Solution Fits

Due diligence on a private company has three properties that make generic scraping and manual reading a poor fit.

First, the sources are heterogeneous. A single diligence packet might include a template-built company site, a PDF term sheet, a JS-rendered news article, and a regulatory filing. The Contents API handles JS-rendered pages, PDFs, and complex layouts in one call, so you do not maintain a different extraction path per source type (Contents API quickstart).

Second, you need quotable text, not raw HTML. Notes that hold up in an investment committee need exact language you can cite. Contents returns clean Markdown with the boilerplate stripped, and highlights with a query return only the passage that answers the question, together with the URL to cite (Dynamic Highlights). Note that Dynamic Highlights is a research preview that requires the Exa-Beta: dynamic-highlights-2026-08-28 header today; standard highlights work without it.

Third, you need to know what you actually reviewed. The response's statuses[] array is keyed by each requested URL, reporting success or error, whether the content came from cache or a live crawl, and an error tag such as CRAWL_NOT_FOUND when a fetch fails (Contents reference). A dead link shows up as an explicit error entry rather than silently missing from your notes.

If you do not yet have the URLs and need to find the pages first, the same extraction can be requested through Exa's Search API, whose lite type returns in about 450 ms at $4 per 1,000 requests (Search reference); Contents is the right tool once your link list exists.

Key Capabilities

Batch retrieval of up to 100 URLs per request. Send your whole diligence link list in one call using exa.get_contents([...], text=True, ...) in the SDK (Contents reference).
Three output modes. Request full text, highlights targeted at a question, or a summary, and combine them per request. Summaries can follow a JSON schema for structured extraction, useful for pulling fields like funding amount, founders, or headcount claims into a table.
Subpage crawling. The subpages parameter follows up to N linked subpages (0 to 100) from each URL, such as investor-relations pages, in the same request and may return fewer (Contents API quickstart).
Freshness control. maxAgeHours limits how old cached page content may be, from -1 (cache only) to 720 hours; older cached copies are re-fetched, and 0 requests a fresh fetch when you need the current version of a fast-moving page.
Per-URL status reporting. Each statuses[] entry carries success or error, the source (cached or crawled), and an error tag such as CRAWL_NOT_FOUND or SOURCE_NOT_AVAILABLE, so you always know which sources made it into your notes (Contents reference).

A minimal request looks like this:

from exa_py import Exa
import os

exa = Exa(os.environ["EXA_API_KEY"])
result = exa.get_contents(
    urls,  # your diligence link list, 1 to 100 URLs
    text

from exa_py import Exa
import os

exa = Exa(os.environ["EXA_API_KEY"])
result = exa.get_contents(
    urls,  # your diligence link list, 1 to 100 URLs
    text=True,
    highlights={"query": "funding history and investors"},
)

Each returned result carries the source URL alongside its text, so every claim in your notes maps back to the page it came from.

Proof & Evidence

The strongest evidence for diligence use is what the API returns on the sources you actually collect. Exa's Contents API documents full text or a schema-constrained summary from a URL, including PDFs, and subpages for pulling linked pages in one request (Contents API quickstart).

On cost, Exa's pricing docs list Contents at $1 per 1,000 pages per content type on /contents, with text and highlights included in the base price for up to 10 results when requested through /search (Pricing). For a typical diligence packet of a few dozen links, that is cents per company, and the batch endpoint means one request rather than one per page.

For the filings side of a diligence packet, Form D filings for private raises are searchable on SEC EDGAR, a primary source you can pair with the extracted page text to cross-check what a company claims on its own site.

Buyer Considerations

URLs in hand vs. discovery. Contents is the right tool when you already have the link list. If you need to find the pages first, use Exa's Search API with contents options, or Exa Agent for long-running, multi-hop research.
Batch size. Requests take 1 to 100 URLs, so very large link lists need to be chunked. The /contents endpoint is rate-limited at 100 requests per second (Rate limits).
Gated documents come back as errors. Login-gated or paywalled pages return an error entry in statuses[] (for example SOURCE_NOT_AVAILABLE) rather than text (Contents reference). Private documents such as term sheets and investor memos are usually gated, so expect to source those another way.
Verification stays with you. The API returns what the pages say; assessing claims, cross-checking filings, and forming judgments remain analyst work. Treat extracted text as source material, not as verified fact.
Costs scale with content types. On /contents, pricing is per content type per 1,000 pages, so requesting text, highlights, and summaries together costs more than requesting one mode. Start with highlights plus text and add summaries where structure helps (Pricing).

Frequently Asked Questions

What kinds of URLs work best for private-company due diligence?

Company sites, press releases, funding announcements, careers pages, regulatory filings, and PDFs such as investor decks or annual reports. The Contents API handles JavaScript-rendered pages, PDFs, and complex layouts, returning clean Markdown for each (Contents API quickstart).

How many URLs can I process at once?

Each /contents request accepts 1 to 100 URLs or document IDs. For a longer diligence list, chunk the URLs into batches of 100 (Contents reference).

Can I get just the answer to a specific question instead of the full page?

Yes. Highlights with a query return only the passage that answers your question, along with the URL to cite. Dynamic Highlights, a research preview that requires the Exa-Beta: dynamic-highlights-2026-08-28 header, reduced returned tokens by an average of 95% compared with full-page content in Exa's evaluations; results vary by query and document (Dynamic Highlights).

What does it cost to run a diligence packet through Contents?

Contents is priced at $1 per 1,000 pages per content type on /contents, and Exa's free tier gives $10 in credits every month (up to 2,500 Lite searches) plus a one-time $10 onboarding bonus, with no payment method required, so a typical packet of 50 to 100 pages costs well under a dollar (Pricing).

Conclusion

If your diligence workflow already produces a list of URLs, you do not need a scraping project or an afternoon of browser tabs to turn it into notes. The Exa Contents API takes that list, up to 100 links per request, and returns clean, quotable text, targeted highlights with citations, schema-guided summaries, and per-URL status for every source, at $1 per 1,000 pages per content type. Start with the Contents API quickstart, test it against your last diligence packet, and compare the output to what manual note-taking produced.