Does Extend AI preserve code blocks and indentation when converting PDFs to Markdown?
Extend AI flattened the PDF’s two code snippets into ordinary text rather than preserving their line structure and indentation. Code characters also changed, and neither snippet would run as returned.
1 of 1 test case failed
Every test case in this scenario has a result.
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.
Code block — fenced, structure intactFailEvidence
from tarnbeck. stations import Export, reload_window def reload_after_reset(station, reset_at): export = Export. open(station) window = reload_window(reset_at, margin_minutes=45) for block in export.blocks(window): if block. checksum_ok(): continue export.requeue (block, reason="gauge reset") return export. commit(dry_run=False)
Returned Markdown: unfenced code with merged statements and lost indentation
Copied from Output file (Markdown), lines 33–39
Open the Markdown file ↗$ ssh loader-01. tarnbeck. example $ tarnbeck-maint reload-after-reset -- station VQ1 | -- reset-at '2026-01-29T09:15' -- confirm
Copied from Output file (Markdown), line 49
{ "parse_run": { "object": "parse_run", "id": "pr_HbkENnP4XWJ7adkx2mXu3", "file": { "object": "file", "id": "file_ffmzz11xz5imXPxtY0uNJ", "name": "operations_note_DS-OP-07.pdf", "type": "PDF", "parentFileId": null, "metadata": { "pageCount": 3 }, "dataRetention": { "mode": "workspace_default", "status": "available" }, "createdAt": "2026-09-09T12:41:15.258Z", "updatedAt": "2026-09-09T12:41:21.669Z" }, "status": "PROCESSED", "failureReason": null, "failureMessage": null, "metadata": null, "dataRetention": { "mode": "workspace_default", "status": "available" }, "output": { "chunks": [ { "object": "chunk", "id": "chunk_1_XzYQdE", "type": "page", "content": "Tarnbeck Institute of Hydrology · Data Services\n\nReloading a station export after a gauge reset\n\nOperations note DS/OP/07 · Data services · Reviewed 14 January 2026\n\n### Symptoms\n\nThe station's daily total is short and the loader's own log … [2,328 characters in the file]", "metadata": { "pageRange": "{… 2 fields in the file}", "minOcrConfidence": 0.332, "avgOcrConfidence": 0.99 }, "blocks": [ "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}" ] }, { "object": "chunk", "id": "chunk_2_4jBDcQ", "type": "page", "content": "Tarnbeck Institute of Hydrology · Data Services\n\nDS/OP/07\n\nfrom tarnbeck. stations import Export, reload_window\n\ndef reload_after_reset(station, reset_at): export = Export. open(station) window = reload_window(reset_at, margin_minutes=45) … [2,400 characters in the file]", "metadata": { "pageRange": "{… 2 fields in the file}", "minOcrConfidence": 0.492, "avgOcrConfidence": 0.989 }, "blocks": [ "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}" ] }, { "object": "chunk", "id": "chunk_3_BiTzmx", "type": "page", "content": "Tarnbeck Institute of Hydrology · Data Services\n\nDS/OP/07\n\n# If the count is short\n\nA short count means blocks were rejected for a second reason and the reload has not found out what it was. Do not run it again: a second pass requeues the s … [1,939 characters in the file]", "metadata": { "pageRange": "{… 2 fields in the file}", "minOcrConfidence": 0.39, "avgOcrConfidence": 0.992 }, "blocks": [ "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}", "{… 8 fields in the file}" ] } ], "metadata": { "originalMimeType": "application/pdf", "finalMimeType": "application/pdf", "pages": [ { "number": 1, "rotationApplied": 0, "originalPageWidth": 596, "originalPageHeight": 842, "dpi": 150 }, { "number": 2, "rotationApplied": 0, "originalPageWidth": 596, "originalPageHeight": 842, "dpi": 150 }, { "number": 3, "rotationApplied": 0, "originalPageWidth": 596, "originalPageHeight": 842, "dpi": 150 } ] } }, "outputUrl": null, "metrics": { "processingTimeMs": 4906, "pageCount": 3 }, "config": { "engine": "parse_performance", "engineVersion": "2.0.0" }, "batchId": null, "usage": { "credits": 6, "totalCredits": 6, "breakdown": [ { "object": "parse_run", "id": "pr_HbkENnP4XWJ7adkx2mXu3", "credits": 6, "charges": ["{… 4 fields in the file}"] } ] } }, "file": { "object": "file", "id": "file_ffmzz11xz5imXPxtY0uNJ", "name": "operations_note_DS-OP-07.pdf", "type": "PDF", "parentFileId": null, "presignedUrl": null, "contents": null, "metadata": {}, "dataRetention": { "mode": "workspace_default", "status": "available" }, "createdAt": "2026-09-09T12:41:15.258Z", "updatedAt": "2026-09-09T12:41:15.282Z" }, "figure_assets": [] }
The whole file (72 KB) is too large to print here; open it from its link below.
Open the JSON file ↗| ✗ | Expected: Emit the code as preformatted or fenced blocks, preserving line breaks and the source’s indentation.Found: Source indentation of 4, 8 and 12 spaces became flush-left text; statements merged and characters changed. Neither snippet was fenced or runnable; loop and conditional nesting was unrecoverable. |
The default pipeline completed the conversion of the full PDF without an error, yet treated genuine code as prose. It did not preserve the required code blocks, line breaks or indentation.
Headings retained. The “Running it” section title remained a Markdown heading.
Headers and footers retained. The Markdown included a page-number tag for each page, alongside the running headers and footers.
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.
How this scenario is graded
How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.
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.
- 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.
| Level | Name | Scope |
|---|---|---|
| Benchmark | Converting a complex PDF into clean Markdown with a hosted API → | 12 scenarios · 13 tools |
| Capability | Code Extraction → | |
| Scenario | A document containing a code block → | |
| Tool | Extend AI → |
Global scenario definition → · Global capability definition → · Extend AI product page →
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.
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.
You run the same kind of test against your own setup and get the same behaviour.
Agree →Yours behaves differently. Tell us what you got, with a screenshot if you have one.
Disagree →Something here is wrong — a reference value, a transcription, a grade.
Report an issue →Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.
Vendor notice →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.
File fingerprints (SHA-256)
A fingerprint identifies the exact file used for this result.