AI Demos Research — the structured-intelligence platform. Every verdict on these pages opens to the execution behind it.
Graded 19 November 2026

Can Extend AI save a corrected field value to the stored record?

Extend AI saved a corrected receipt number to the stored record: after RHS-70234 was changed to RHS-99999 on its review screen, saved and the run reloaded, the stored record read RHS-99999. The subtotal, tax and total due stayed at 85, 5.1 and 90.1.

1 of 1 test case passed

Every test case in this scenario has a result.

Pass rate100%1 of 1 gradable
Coverage1 of 1test case with a result
1PassDid everything it was expected to do.
0FailNo test case failed.
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.

Corrected value reaches the stored recordPassEvidence
What we sent

Input 1 · Edit before saving (screenshot): Typed RHS-99999 over the receipt number field on the saved run's review screen, before saving.

Input 2 · Same edit on a second saved run (recording): Typed RHS-99999 over the receipt number field on a second saved run of the same receipt, with Save corrections offered.

InputDocument (PDF)
Input PDF: the one-page service receipt, with receipt number RHS-70234 and its totals
Open the PDF ↗
What the tool returned
Textthe whole file · copied lines 4–18 marked
Extend AI run exr_o7A2RVk90ZLc9JJjOur8g (Baseline_receipt.pdf), Runs > "Actual vs expected (diff)" > JSON view,
read after pressing "Save corrections" and reloading the run page (2026-10-06, IST).

Actual output (extracted in this run — unchanged):
{
  "tax": 5.1,
  "subtotal": 85,
  "total_due": 90.1,
  "receipt_number": "RHS-70234"
}

Expected output (the human-reviewed / stored record — includes the correction; run is badged "Reviewed"):
{
  "tax": 5.1,
  "subtotal": 85,
  "total_due": 90.1,
  "receipt_number": "RHS-99999"
}

Transcript: both sides of the after-save comparison, the original extraction and the stored record

Copied from Captured file (JSON), lines 4–18

Open the JSON file ↗
Expected vs. Found
✓Expected: The stored record holds the corrected value, and nothing else changed.Found: After saving and reloading, the stored record read receipt number RHS-99999 in place of RHS-70234; subtotal 85, tax 5.1 and total due 90.1 were identical before and after.
Supporting proof
Proof 1Screenshot
Screenshot after save and reload: the original extraction beside the stored reviewed record, in JSON
Open original ↗
Proof 2Screenshot
Screenshot after save and reload: each field's original and reviewed values, side by side
Open original ↗
Proof 3Screenshot
Screenshot before saving: the receipt number typed over and badged Corrected
Open original ↗
Proof 4Screen recording
Screen recording: the edit, save and reload on a second saved run of the same receipt
Key moments
Open original ↗
Proof 5Captured file (JSON)
Original extraction output: the four fields as first returned, before any correction
Cited values2 values from this file
value
{ "tax": 5.1, "subtotal": 85, "total_due": 90.1, "receipt_number": "RHS-70234" }
metadata › receipt_number
{ "citations": [ { "page": 1, "fileId": "file_eWxLbOliICKuI7XAugIcq", "polygon": [ { "x": 49.5792, "y": 194.29919999999998 }, { "x": 107.7624, "y": 194.2776 }, { "x": 107.7624, "y": 205.1064 }, { "x": 49.5792, "y": 205.1352 } ], "pageWidth": 594, "pageHeight": 841.0032, "referenceText": "RHS-70234" } ], "ocrConfidence": 0.994 }
Show the whole file (2 KB)
JSON
{
  "value": {
    "tax": 5.1,
    "subtotal": 85,
    "total_due": 90.1,
    "receipt_number": "RHS-70234"
  },
  "metadata": {
    "tax": {
      "citations": [
        {
          "page": 1,
          "fileId": "file_eWxLbOliICKuI7XAugIcq",
          "polygon": [
            { "x": 308.9448, "y": 312.0912 },
            { "x": 348.3, "y": 312.5088 },
            { "x": 348.1848, "y": 323.568 },
            { "x": 308.916, "y": 323.14320000000004 }
          ],
          "pageWidth": 594,
          "pageHeight": 841.0032,
          "referenceText": "Tax (6%)"
        }
      ],
      "ocrConfidence": 0.997
    },
    "subtotal": {
      "citations": [
        {
          "page": 1,
          "fileId": "file_eWxLbOliICKuI7XAugIcq",
          "polygon": [
            { "x": 309.132, "y": 294.35760000000005 },
            { "x": 344.664, "y": 294.37919999999997 },
            { "x": 344.65680000000003, "y": 304.1784 },
            { "x": 309.1248, "y": 304.1496 }
          ],
          "pageWidth": 594,
          "pageHeight": 841.0032,
          "referenceText": "Subtotal"
        }
      ],
      "ocrConfidence": 0.989
    },
    "total_due": {
      "citations": [
        {
          "page": 1,
          "fileId": "file_eWxLbOliICKuI7XAugIcq",
          "polygon": [
            { "x": 308.9664, "y": 336.564 },
            { "x": 338.35679999999996, "y": 336.62879999999996 },
            { "x": 338.328, "y": 348.0552 },
            { "x": 308.9376, "y": 347.99039999999997 }
          ],
          "pageWidth": 594,
          "pageHeight": 841.0032,
          "referenceText": "Total"
        }
      ],
      "ocrConfidence": 0.996
    },
    "receipt_number": {
      "citations": [
        {
          "page": 1,
          "fileId": "file_eWxLbOliICKuI7XAugIcq",
          "polygon": [
            { "x": 49.5792, "y": 194.29919999999998 },
            { "x": 107.7624, "y": 194.2776 },
            { "x": 107.7624, "y": 205.1064 },
            { "x": 49.5792, "y": 205.1352 }
          ],
          "pageWidth": 594,
          "pageHeight": 841.0032,
          "referenceText": "RHS-70234"
        }
      ],
      "ocrConfidence": 0.994
    }
  }
}
Open the JSON file ↗
Why this result

The corrected value had to reach the stored record, with nothing else changing. After saving and reloading, the stored record holds the new receipt number and the other three fields are identical, so the edit did not stay only on screen, was not lost and altered no other field.

Also observed in this test case

Original extraction kept alongside. The run keeps the original extraction, receipt number RHS-70234, as its actual output and stores the correction beside it as the human-reviewed value, with the run badged Reviewed.

Corrections flagged before saving. Typing over the receipt number badged the field Corrected at once; an Unsaved changes notice and a Save corrections control appeared, and a Corrections saved message confirmed the save.

Tested by Rugved Nichite · evidence dated 6 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
Extend AI
Build or version
Not exposed by the hosted product
Surface
Web app
Set up before the run
The baseline service receipt and its schema were used. On the saved run's own review screen, the correctly extracted receipt number was changed to another valid value, and the record was then read back.
Review screen
the saved-run review screen, not the Studio Playground
Field corrected
receipt number
Tested
6 October 2026 · Rugved Nichite

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
CapabilityReview and Correction →
ScenarioCorrect a field value →
ToolExtend AI →

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
20 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/extend-ai/correct-a-field-value · 1 pass · 0 fail · coverage 1/1 · graded 2026-11-19

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 · 127 KB
Proof 2 · Screenshot ↗image/png · 187 KB
Proof 3 · Screenshot ↗image/png · 54 KB
Captured file (JSON) ↗text/plain; charset=utf-8 · 1 KB
Proof 5 · Captured file (JSON) ↗application/json · 2 KB
File fingerprints (SHA-256)

A fingerprint identifies the exact file used for this result.

Input · Document (PDF)1f464ae863aadb1682805b8ee3fdd8b368a2ddd4e0f602e1cb194640b9253302
Proof 1 · Screenshot6af2aa7bb7144e801bdcf9904ba37e90c0f6158d0e4b189da017b1e067f5761f
Proof 2 · Screenshot78e13b6637031c83d5b32d6108d158988b260d98f159bdfef9cbd4b5e7110339
Proof 3 · Screenshot92489b423cf6015ad93083e44deb60daa62e9fcfd6f2136300502ae76a627219
Proof 4 · Screen recording4c6a2112e187b65165eaff07ed9674770661b3a06fdc53baa483f2fce98bd41d
Captured file (JSON)cf5ac039df369e65ea9c3c40bba12c37243ca922c08d3c50b3845e71a6955c2a
Proof 5 · Captured file (JSON)a8fe307a48b8463564205116ae2769a4ceb1031b7eaff669dd31f106124c7867