// see it in action

See exactly what you get back.

Three real-world documents parsed end-to-end by the production pipeline — pick one and inspect the raw output next to the source. No signup, no credit card.

paper-scan.pdfOutput: markdownModel: PaddleOCR-VLConfidence: 0.97Sample documents are fictional; this is the real DocUnwrap output format.

Parse result

Page 1

The DocUnwrap Parsing Pipeline: Routing, Confidence, and Page-Level Review

DocUnwrap Engineering Team

DocUnwrap Technical Report Series · TR-26-114 · September 2026

Abstract

This report describes the parsing pipeline behind DocUnwrap. Every page is treated as an independent unit: a router inspects the page image, assigns a model matched to its degradation profile, and attaches a confidence score to the result. Pages that fall below the confidence threshold are escalated to a stronger model before the caller ever sees them. We summarize the default routing policy, the page-level review flags, and the operational rules that keep the pipeline predictable under load.

1. Why Routing

Not all pages are equally hard. A text-native PDF can be read from its embedded text layer in milliseconds, while a skewed phone photo may need a full vision-language pass. A single-model stack therefore overpays for easy pages and still underperforms on hard ones. DocUnwrap treats routing as the first-class decision, and every downstream field hangs off it.

2. Pipeline Stages

  1. Intake — documents arrive through the API or the console; pages get stable identifiers.
  2. Parse — each page is routed per the policy below and parsed into content blocks.
  3. Assemble — page results merge into one Markdown document or a per-page JSON record.

2.1 Routing Policy

Table 1: Default routing policy

Page conditionRouteNotes
Text-native PDFText layerNo GPU inference, near-instant
Clean scanpaddleocr-vlHigh-accuracy layout model
Photo or heavy skewteleocrDegradation-tuned re-parse
Confidence below barglm-ocrEscalated before delivery

3. What Callers See

Callers receive clean Markdown with the heading hierarchy, lists and tables intact, or a per-page JSON record that carries the page text, its confidence, the model that produced it and a needsReview flag — everything a review queue or a RAG indexer needs to act on its own.

Parse your own documents

Create an account and parse your first 100 pages free — upload a PDF or call the API.

Start free