Innovation & Tech

You.com’s Extract API Brings Full-PDF Parsing to Developers

You.com’s Extract API parses full PDFs into Markdown or HTML. Lite costs $0.90 per 1,000 URLs; standard reaches $13.90 per 1,000 typical 26-page files and slows responses.

By Daniel Okafor

4 min read

You.com Launches Extract API for Document Retrieval and PDF Parsing
You.com Launches Extract API for Document Retrieval and PDF ParsingGoumbik / pixabay

What's News

  • You.com launched the Extract API on November 17, 2026.
  • You.com reported Extract standard scores of 74.9 for tables and 76.6 for charts, versus 22.8 and 0.9 for Exa Contents.
  • You.com reported median latency of 11.0 seconds for standard and 2.6 seconds for lite, versus 1.3 seconds for Exa and 1.8 seconds for Parallel.
  • Lite costs $0.90 per 1,000 URLs; standard reaches $13.90 per 1,000 typical 26-page documents.
  • You.com estimates $100 in work-email credit covers about 7,000 typical standard-mode PDFs.

By Daniel Okafor

You.com launched the Extract API on November 17, 2026, and puts standard PDF parsing at $13.90 per 1,000 typical 26-page documents. You.com said the new endpoint fetches a web page or PDF and returns its contents as Markdown or HTML.

Developers can call it after a search for a filing, contract or research paper. They can also call it directly when they already know the URL. You.com reported that its default PDF mode scored at least 80% higher than Exa Contents and Parallel Extract in a five-dimension evaluation.

How does Extract retrieve documents?

The API reference defines three retrieval settings, each with a specific cache and live-fetch behavior.

  • blend is the default. It uses cached content when available and crawls the page live otherwise.
  • fetch always crawls the page live.
  • cache returns stored content only. It omits the content field when no copy exists.

Requests require an API key. Each response identifies whether the material came from a cache or a live fetch.

The endpoint has a 30-second end-to-end timeout by default. Operators can set the limit from 2 to 100 seconds. The API returns a 504 error after a timeout.

What separates standard from lite?

PDF processing runs through a separate parse_effort setting. Standard is the default. Lite extracts only text embedded in each page.

  • standard sends complex pages, including charts and images, through an advanced model. You.com says it targets tables, charts, forms and scanned pages.
  • lite suits digitally created, text-based documents such as contracts, according to You.com.

Neither setting changes web-page output or pricing, You.com said.

How strong are the parsing results?

The evaluation covered 3,782 pages across 12 ParseBench-2 sets. You.com ran both Extract modes alongside Exa Contents and Parallel Extract. Each service received the same URL. ParseBench’s grader scored the outputs, and a service that failed to return a document received zero.

You.com reported the widest advantages in tables and charts. Extract standard scored 74.9 for tables and 76.6 for charts. Exa Contents scored 22.8 and 0.9 in those categories. Parallel scored 0.1 and zero.

Speed moved the other way. You.com reported median latency of 11.0 seconds for standard and 2.6 seconds for lite. Its comparison put Exa at 1.3 seconds and Parallel at 1.8 seconds. You.com noted that Parallel’s minimum cache age of 600 seconds means some responses may have been cached.

The comparison does not establish a full ParseBench-2 ranking. You.com excluded the coordinate-based layout evaluation because Extract returns parsed content rather than page-element coordinates.

LlamaIndex’s public leaderboard covers 56 methods, 4,317 pages and 13 datasets across six dimensions. It calculates the overall score as the plain mean of those dimensions. You.com said the 80% comparison describes its scoped test, not the full ParseBench-2 ranking.

What does Extract cost?

Lite has a flat price of $0.90 per 1,000 URLs. Standard uses that URL rate and adds $0.50 per 1,000 pages when You.com identifies the fetched file as a PDF.

Based on the 26-page median across the billions of PDFs it has processed, You.com puts the cost of 1,000 typical documents at $13.90 when using standard. Lite remains $0.90.

You.com’s comparison lists Exa and Parallel at $1.00 per 1,000 URLs regardless of document length. You.com described standard as carrying a substantial multi-page premium. It said lite undercuts those services for text-based PDFs.

You.com does not bill requests that fail on its side. Its error responses distinguish blocked URLs, failed requests and pages without content. Developers can use those signals to decide whether to retry.

Zero data retention is available for sensitive documents at no additional charge. Customers must contact You.com to enable it.

What happens to the older Contents API?

Developers can use Extract on web pages through You.com’s official MCP server for free. It requires neither an API key nor authentication.

Direct API sign-ups include $10 in credit, or $100 with a work email. You.com estimates that the larger amount covers about 7,000 typical PDFs on standard.

The older Contents API remains supported for existing users. It is no longer in active development. Deeper PDF parsing and parse_effort remain exclusive to Extract.

Which mode fits a retrieval pipeline?

The choice depends on where the answer sits. Lite preserves a flat per-URL cost and faster response when a PDF already contains embedded text. Standard takes longer and costs more, but You.com’s results show much stronger table and chart parsing.

The next procurement test is straightforward: whether those parsing gains justify $13.90 per 1,000 typical documents and an 11.0-second median wait. On You.com’s figures, lite remains the cheaper, faster route when embedded text already holds the answer.

Source: unite.ai

Share this article:

More from Daniel Okafor

Daniel Okafor

Show full bio

Correspondent covering money, funding and market shifts around entrepreneurship, startups and venture funding at Business Bearings.

2,282 articles

Related articles

« Previous articleNext article »