hugepdf.io

Any PDF. Any size. Per-page processing. Agent-first.

What It Does

hugepdf.io extracts PDFs page-by-page with a triple extraction pipeline, then reconciles the outputs into one structured result. It is built for agents that need reliable extraction on complex layouts.

Triple Extraction

Each page is processed through:

1. Python text extraction
2. JavaScript text extraction
3. Vision extraction on page images

A reconciliation prompt merges these views so single-parser blind spots are reduced.

Pricing

Pricing is per page and explicit in API responses. Default target rate: $0.03 per page. Accepted tokens are SUI and USDC on Sui.

How To Try It

Submit a PDF with POST /api/process, then call GET /job/{job_id}/dry-run?token={output_token} to evaluate sample extraction quality before paying for full processing.

Agent-Native Payments

No API keys. No account required. Payment on-chain is authentication and identity.

Machine Spec

For agent discovery and integration details, use /llms.txt.