AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
Graded 17 November 2026

Does Reducto distinguish a document answer from extracted data when the fact is outside the schema?

Reducto gives no way to ask a question about a document, so it cannot say whether an answer came from the extracted data or the document itself. After one Parse request the technician printed on the receipt appeared among the parsed blocks; after one Extract request only the four schema fields came back.

1 of 1 test case failed

Every test case in this scenario has a result.

Pass rate0%0 of 1 gradable
Coverage1 of 1test case with a result
0PassNo test case passed.
1FailMissed at least one thing it was expected to do; the reason and proof are with each test case below.
0Not gradableEvery result here could be graded.
0UntestedEvery test case in this scenario has a result.
The pass rate is a summary. The evidence is the test case below: what we sent, what we checked, what the tool returned, and the proof.

The test case

Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on this test case.

Answer outside the schema, attributedFailEvidence
What we sent

In Reducto Studio, we ran the service receipt once through Parse and once through Extract, with a schema for the receipt number, subtotal, tax and total due, and looked on both result screens for a place to ask who the receipt's technician was.

InputDocument (PDF)
Input PDF: the one-page service receipt that names the technician
Open the PDF ↗
Extraction schema
receipt_number, subtotal, tax, total_due
What the tool returned

The tool's response has no text to quote: it is what the evidence under Proofs shows.

Proof 2 · Screenshot

Proof 1 · Screenshot

Expected vs. Found
✗Expected: It tells the two apart: it says the fact is not in the extracted data, or it answers and says the answer came from the document.Found: Neither the Parse nor the Extract screen offers a question box, chat or search field. Parse showed the technician; Extract returned only the four schema fields.
Supporting proof
Proof 1Screenshot
Parse screenshot: the technician in the parsed content, with no question input
Open original ↗
Proof 2Screenshot
Extract screenshot: four extracted fields without the technician, and no question input
Open original ↗
Why this result

Telling the two apart needs a way to ask, and Reducto offers none: neither result screen has a question surface, and its documented operations, Parse, Extract, Split and Edit, include none that answers questions. The failure is the product's, not the run's.

Tested by Shivohum Pandita · evidence dated 29 October 2026

Configuration and setup

How this tool was set up for the run and what the test needed in place. Each row is a fact from the run's records; a fact the records do not hold is left out, not guessed.

Software that produced the output
Reducto
Build or version
Not exposed by the hosted product
Plan or tier
Standard
Model
Extract v4 (the default extract model) and parse model r-1 (Preview)
Surface
Web app
Set up before the run
The receipt had to state a fact plainly that the schema left out: here the technician named on the receipt, outside a schema of receipt number, subtotal, tax and total due. An answer from the document would be correct.
Citations
on
Tested
29 October 2026 · Shivohum Pandita

How this scenario is graded

How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.

How results are decided

Each test case gets one result per tool: Pass, Fail or Not gradable. A test case we haven't run yet shows Untested. There are no partial results.

The rules
  • Pass: the tool did everything the test expected, and nothing it said contradicts the correct answer.
  • Fail: at least one expected behaviour clearly didn't happen; the test case's result says which one and quotes the tool.
  • Not gradable: our evidence couldn't settle the outcome (for example a record we needed is missing). It is never counted as a fail, and the test case's result says what's missing.

Where this sits in the benchmark

This page is one cell of a larger study: one tool, one scenario. Only this benchmark's frame appears here.

LevelNameScope
BenchmarkStructured Document Extraction →19 scenarios · 10 tools
CapabilityQuerying →
ScenarioThe answer was never in the schema →
ToolReducto →

History of this result

What has happened to this result since it was first published. Runs and grades are never overwritten: a retest or a re-grade publishes a new result and keeps the earlier one readable.

from the publication record
18 November 2026First publishedStructured Document Extraction v1

Act on this result

Nothing filed here edits the run or the grade. A challenge opens a review, and a review can produce a new run or a re-grade — which becomes the current result and leaves this one in the history.

This matches what I see

You run the same kind of test against your own setup and get the same behaviour.

Agree →
This does not match

Yours behaves differently. Tell us what you got, with a screenshot if you have one.

Disagree →
Point out an issue

Something here is wrong — a reference value, a transcription, a grade.

Report an issue →
Request a retest

On a newer build, a larger dataset, or your own setup.

Request a retest →
We have fixed this

Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.

Vendor notice →
Filed against this evidenceNothing yet. Challenges, counter-evidence and fix notices appear here with their outcome, and stay on the page after they are resolved.
Cite this result
aidemos.com/benchmarks/structured-document-extraction/results/reducto/the-answer-was-never-in-the-schema · 0 pass · 1 fail · coverage 1/1 · graded 2026-11-17

The same record is available as structured data through the AI Demos MCP server, with the counts, the coverage and every per-test-case reason carried as fields.

Verify the proof files

These files support this result. Open a file to inspect the original evidence.

Input · Document (PDF) ↗application/pdf · 7 KB
Proof 1 · Screenshot ↗image/png · 229 KB
Proof 2 · Screenshot ↗image/png · 314 KB
File fingerprints (SHA-256)

A fingerprint identifies the exact file used for this result.

Input · Document (PDF)1f464ae863aadb1682805b8ee3fdd8b368a2ddd4e0f602e1cb194640b9253302
Proof 1 · Screenshota2dfeeca3d7e54084a610b574a1eec580b7a2810ddafa464e981d0e6afb4e89b
Proof 2 · Screenshot6d18cc5b6b2b72cf0251d10d81dec50ac0eb3558f5f51e460083cb6b55b7407c