Tool in benchmark · Version 1

Unstructured in Converting a complex PDF into clean Markdown with a hosted API

Scenario-level performance from current published Results.

1Passed
4Failed
7No published Result yet

5 of 12 scenarios published

CapabilityScenariosResult
Figures & Charts0 pass · 1 fail · 1 of 2 published
Heading & Section Structure0 pass · 1 fail · 1 of 2 published
Reading Order & Layout0 pass · 1 fail · 1 of 1 published
Scanned Document OCR1 pass · 0 fail · 1 of 2 published
Text Fidelity0 pass · 1 fail · 1 of 1 published
Code Extraction0 of 1 published
Table Extraction0 of 2 published
PassedFailedPublished, different test setNo published Result yet

Where Unstructured failed

4 of 5 graded scenarios. Open a Result to see the conversation and the evidence.

Figures & Charts

A document with a data chart

Unstructured's default conversion did not preserve the data chart as an image in Markdown. It supplied a text description, while the plan on page 2 retained its image.

View Result →
Heading & Section Structure

A document with styled headings and subheadings

Unstructured returned the document's headings at a single Markdown level, so the source's three-tier nesting was lost. Of the 26 section headings in the source, 11 came back as headings, all at one level, and the other 15 ran into the opening words of plain paragraphs. The raw response gave no heading a level or a parent heading.

View Result →
Reading Order & Layout

A page laid out in multiple columns

Unstructured read the columns of a two-column bulletin in turn except for one section: on page 3, the chapels survey starts at the foot of the left column and continues at the top of the right. Unstructured output the continuation before the Membership section, and the survey's heading and opening paragraph only at the very end, after Notices. The text therefore does not come out in natural reading order.

View Result →
Text Fidelity

An ordinary digital text document

Unstructured fails to keep an ordinary digital text document complete and unchanged. The returned Markdown matches the body text after page furniture is removed, but all seven page footers become headings and four body sentences are split by running-header words.

View Result →

How Unstructured performed

Open a capability to explore its scenarios. Each row reports the test set in its published Result.

Figures & Charts1 fail1 no Result yet
ScenarioPublished outcomesTest coverageResult
A document with a data chart
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Unstructured's default conversion did not preserve the data chart as an image in Markdown. It supplied a text description, while the plan on page 2 retained its image.

1/1 assessed1/1 gradablePublished test setView Result →
A document with figures and captionsNo published result
Heading & Section Structure1 fail1 no Result yet
ScenarioPublished outcomesTest coverageResult
A document with styled headings and subheadings
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Unstructured returned the document's headings at a single Markdown level, so the source's three-tier nesting was lost. Of the 26 section headings in the source, 11 came back as headings, all at one level, and the other 15 ran into the opening words of plain paragraphs. The raw response gave no heading a level or a parent heading.

1/1 assessed1/1 gradablePublished test setView Result →
Footnotes at the bottom of the pageNo published result
Reading Order & Layout1 fail
ScenarioPublished outcomesTest coverageResult
A page laid out in multiple columns
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Unstructured read the columns of a two-column bulletin in turn except for one section: on page 3, the chapels survey starts at the foot of the left column and continues at the top of the right. Unstructured output the continuation before the Membership section, and the survey's heading and opening paragraph only at the very end, after Notices. The text therefore does not come out in natural reading order.

1/1 assessed1/1 gradablePublished test setView Result →
Scanned Document OCR1 pass1 no Result yet
ScenarioPublished outcomesTest coverageResult
A cleanly scanned document
1 Pass0 Fail0 Not gradable
1 of 1 test case passed
What happened

Unstructured accurately recovered text from a clean, image-only scan. It preserved the certificate’s measurements, units and dates without adding text.

1/1 assessed1/1 gradablePublished test setView Result →
A document mixing digital and scanned pagesNo published result
Text Fidelity1 fail
ScenarioPublished outcomesTest coverageResult
An ordinary digital text document
0 Pass1 Fail0 Not gradable
1 of 1 test case failed
What happened

Unstructured fails to keep an ordinary digital text document complete and unchanged. The returned Markdown matches the body text after page furniture is removed, but all seven page footers become headings and four body sentences are split by running-header words.

1/1 assessed1/1 gradablePublished test setView Result →
Code Extraction1 no Result yet
ScenarioPublished outcomesTest coverageResult
A document containing a code blockNo published result
Equations & Mathematical Notation1 no Result yet
ScenarioPublished outcomesTest coverageResult
A document containing mathematical equationsNo published result
Table Extraction2 no Result yet
ScenarioPublished outcomesTest coverageResult
A simple, clearly formatted tableNo published result
A table that continues across a page breakNo published result

Reading these Results

Published evidence and test coverage answer different questions.

Publication availability

Which scenarios have a Result?

A published Result is public evidence for this tool on one scenario. “No published result” does not say whether testing has taken place.

Test coverage

What does each Result cover?

Assessed includes Pass, Fail and Not gradable. Gradable includes Pass and Fail. Both use the pinned test count in that published Result.

Scenario scope

Out of 12 scenarios

The summary counts each of the 12 scenarios once. A scenario fails when any of its published test cases failed. “No published Result yet” means nothing is public for that scenario; it does not say whether Unstructured was tested on it.

Unstructured in Converting a complex PDF into clean Markdown with a hosted API | AI Demos