{"observation":{"id":"b83ea780-c133-4101-8a90-10799b20ab97","tool":"llamaparse","tool_name":"LlamaParse","criterion":"table-preservation","criterion_name":"Table Preservation","criterion_definition":"Preserves complex table structures, including rows, columns, multi-row headers, and merged-cell relationships in markdown.","criterion_evidence_type":"transformation","criterion_rank_role":"decisive","criterion_rank_role_reason":"Accurate Markdown conversion of complex PDFs depends on keeping table structure intact, not flattening it into plain text. (3 of 3 judges)","scenario":"financial-report-table-heavy","scenario_name":"Financial Report - Table Heavy","group_tag":"financial-report-table-heavy","scenario_description":"A table-heavy corporate financial report used to test extraction of dense, hierarchical financial statements with grouped columns, multi-row headers, segment-reporting tables, and narrative disclosures.","modality":"pdf","input_text":null,"input_artifact_refs":[{"alt":null,"url":"https://cdn.futuresmart.ai/public/aidemos/fba42ade5adb4814a3381dd4487a3910.pdf?v=1","role":"input","filename":"doc.pdf"}],"stresses":["Multi-page financial statement extraction","Hierarchical table reconstruction","Grouped columns and multi-row headers","Reading order in a report with mixed narrative and tables","Document structure retention","Markdown usability"],"verdict":"worked","score":null,"score_total":null,"note":"Preserves a multi-level segment-results table with grouped headers and year-over-year columns, keeping the hierarchy legible in the extracted output.","evidence_state":"verified","source":null,"artifacts":[{"url":"https://cdn.futuresmart.ai/public/aidemos/808bccab0f4e4188ad2e46b4a0124763.png?v=1","role":"input","alt":null},{"url":"https://cdn.futuresmart.ai/public/aidemos/257fe280b9c74cf1b14a8ec50f7fd151.png?v=1","role":"output","alt":null}],"run_id":"6e3160de-fe46-4b45-b071-72560b5c5d0e","study_title":"Convert a Complex PDF into Clean Markdown with an API","study_kind":"generation","research_task":"86b9h7t37","tested_at":null,"completeness":"input-and-output","input":{"state":"files","text":null,"files":[{"url":"https://cdn.futuresmart.ai/public/aidemos/fba42ade5adb4814a3381dd4487a3910.pdf?v=1","filename":"doc.pdf","alt":"Financial Report - Table Heavy","role":"input"}],"modality":"pdf","stresses":["Multi-page financial statement extraction","Hierarchical table reconstruction","Grouped columns and multi-row headers","Reading order in a report with mixed narrative and tables","Document structure retention","Markdown usability"]},"tool_page_slug":"llamaparse","tool_url":"https://aidemos.com/tools/llamaparse","permalink":"https://aidemos.com/evidence/b83ea780-c133-4101-8a90-10799b20ab97","api_url":"https://ai.aidemos.com/v1/observations/b83ea780-c133-4101-8a90-10799b20ab97"},"peers":[{"id":"0e673751-dc3c-49ef-8cca-67cd82fab54f","tool":"adobe-api","tool_name":"Adobe API","verdict":"worked","score":null,"score_total":null,"note":"Reconstructs the quarterly balance-sheet table with its major asset rows and the two reporting dates, preserving the core balance-sheet layout.","artifact_count":2,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/85b9315b2b0d4b2c911aa582e9a51372.png?v=1","evidence_url":"https://aidemos.com/evidence/0e673751-dc3c-49ef-8cca-67cd82fab54f"},{"id":"97367dd9-249f-4997-95e4-14cc616e9713","tool":"extend-ai","tool_name":"Extend AI","verdict":"worked","score":null,"score_total":null,"note":"Preserves grouped-column structure in the financial tables, including the multilevel segment matrix and the balance-sheet comparison layout.","artifact_count":3,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/7e4a0f994215484c920ba868f35fcef4.png?v=1","evidence_url":"https://aidemos.com/evidence/97367dd9-249f-4997-95e4-14cc616e9713"},{"id":"ff397516-558d-48d7-bbab-039a3f745539","tool":"landing-ai","tool_name":"Landing AI","verdict":"mixed","score":null,"score_total":null,"note":"Recovers the numeric values from a complex table but merges nested header levels into a single cell, weakening the semantic relationships of the table structure.","artifact_count":2,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/42af0b6a8200417588083a90fa79bbc7.png?v=1","evidence_url":"https://aidemos.com/evidence/ff397516-558d-48d7-bbab-039a3f745539"},{"id":"5fe802fc-62a5-4a6a-aa9c-9ae77852cde1","tool":"mistral-ai","tool_name":"Mistral AI","verdict":"failed","score":null,"score_total":null,"note":"Distinct header levels are collapsed into a single cell in the multilevel table, weakening the parent-child relationships that define the column semantics.","artifact_count":3,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/79983e2875fa4de3adacd9d60f5e9dd0.png?v=1","evidence_url":"https://aidemos.com/evidence/5fe802fc-62a5-4a6a-aa9c-9ae77852cde1"},{"id":"a86ffd95-175a-40ec-8670-6791368379df","tool":"nutrient-io","tool_name":"Nutrient.io","verdict":"failed","score":null,"score_total":null,"note":"Loses multi-level header organization in hierarchical tables, so parent-child column relationships are not reconstructed in the markdown output.","artifact_count":4,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/33be110e75ed4e7ca9c753431c9ecf0c.png?v=1","evidence_url":"https://aidemos.com/evidence/a86ffd95-175a-40ec-8670-6791368379df"},{"id":"8c8cffd6-9942-4cc3-95c3-b837dfd045e3","tool":"reducto","tool_name":"Reducto","verdict":"worked","score":null,"score_total":null,"note":"Preserves the balance-sheet hierarchy and values, including Assets > Current assets > Cash and deposits, and also reconstructs the later 9-column segment table with all values correct.","artifact_count":4,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/ae13c7e19e8b412abc1f87bcc21554f3.png?v=1","evidence_url":"https://aidemos.com/evidence/8c8cffd6-9942-4cc3-95c3-b837dfd045e3"},{"id":"63381a32-714a-48c6-b8dc-3ce4fc67279d","tool":"tensorlake","tool_name":"Tensorlake","verdict":"failed","score":null,"score_total":null,"note":"Breaks on a more complex multi-header table by losing header hierarchy and omitting at least one header label, producing an incomplete and structurally incorrect reconstruction.","artifact_count":2,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/bf75efe477c745c68266f27b04959671.png?v=1","evidence_url":"https://aidemos.com/evidence/63381a32-714a-48c6-b8dc-3ce4fc67279d"},{"id":"c66e2b14-a549-4ce7-bbd9-4f0fb50a2057","tool":"upstage-ai","tool_name":"Upstage AI","verdict":"failed","score":null,"score_total":null,"note":"Misaligns column headers with data regions in dense financial tables, producing a structurally inconsistent balance-sheet extraction rather than a clean header-to-row mapping.","artifact_count":4,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/06a0e21888cc4bb187dec34ceb255186.png?v=1","evidence_url":"https://aidemos.com/evidence/c66e2b14-a549-4ce7-bbd9-4f0fb50a2057"}],"other_criteria":[{"id":"c2b3fb59-2c32-4edb-a21b-125cb364f2ea","criterion":"complex-document-handling","criterion_name":"Complex Document Handling","rank_role":"decisive","verdict":"worked","score":null,"score_total":null,"note":"Processes an 18-page table-heavy report end-to-end and reaches SUCCESS in the Results view without manual correction.","artifact_count":2,"evidence_url":"https://aidemos.com/evidence/c2b3fb59-2c32-4edb-a21b-125cb364f2ea"},{"id":"91c495ea-e42f-469b-8dbf-70720d32cb86","criterion":"markdown-quality","criterion_name":"Markdown Quality","rank_role":"decisive","verdict":"mixed","score":null,"score_total":null,"note":"Renders the table of contents as sequential text with page numbers instead of a nested TOC structure, so the markdown is usable but flattened.","artifact_count":1,"evidence_url":"https://aidemos.com/evidence/91c495ea-e42f-469b-8dbf-70720d32cb86"},{"id":"06f3e900-c9d1-4767-bae7-efa3ac0e9779","criterion":"reading-order-structure","criterion_name":"Reading Order & Structure","rank_role":"decisive","verdict":"worked","score":null,"score_total":null,"note":"Preserves the document title, disclaimer, and section headings in a readable hierarchy for the Sumitomo Heavy Industries financial report.","artifact_count":2,"evidence_url":"https://aidemos.com/evidence/06f3e900-c9d1-4767-bae7-efa3ac0e9779"}],"appears_in":[{"page_type":"ranking","slug":"pdf-to-markdown-apis","title":"Best AI Tools to Convert Complex PDFs into Clean Markdown with an API","url":"https://aidemos.com/best/pdf-to-markdown-apis","binding":"run"}],"same_scenario":[{"id":"0e673751-dc3c-49ef-8cca-67cd82fab54f","tool":"adobe-api","tool_name":"Adobe API","verdict":"worked","score":null,"score_total":null,"note":"Reconstructs the quarterly balance-sheet table with its major asset rows and the two reporting dates, preserving the core balance-sheet layout."},{"id":"97367dd9-249f-4997-95e4-14cc616e9713","tool":"extend-ai","tool_name":"Extend AI","verdict":"worked","score":null,"score_total":null,"note":"Preserves grouped-column structure in the financial tables, including the multilevel segment matrix and the balance-sheet comparison layout."},{"id":"ff397516-558d-48d7-bbab-039a3f745539","tool":"landing-ai","tool_name":"Landing AI","verdict":"mixed","score":null,"score_total":null,"note":"Recovers the numeric values from a complex table but merges nested header levels into a single cell, weakening the semantic relationships of the table structure."},{"id":"5fe802fc-62a5-4a6a-aa9c-9ae77852cde1","tool":"mistral-ai","tool_name":"Mistral AI","verdict":"failed","score":null,"score_total":null,"note":"Distinct header levels are collapsed into a single cell in the multilevel table, weakening the parent-child relationships that define the column semantics."},{"id":"a86ffd95-175a-40ec-8670-6791368379df","tool":"nutrient-io","tool_name":"Nutrient.io","verdict":"failed","score":null,"score_total":null,"note":"Loses multi-level header organization in hierarchical tables, so parent-child column relationships are not reconstructed in the markdown output."},{"id":"8c8cffd6-9942-4cc3-95c3-b837dfd045e3","tool":"reducto","tool_name":"Reducto","verdict":"worked","score":null,"score_total":null,"note":"Preserves the balance-sheet hierarchy and values, including Assets > Current assets > Cash and deposits, and also reconstructs the later 9-column segment table with all values correct."},{"id":"63381a32-714a-48c6-b8dc-3ce4fc67279d","tool":"tensorlake","tool_name":"Tensorlake","verdict":"failed","score":null,"score_total":null,"note":"Breaks on a more complex multi-header table by losing header hierarchy and omitting at least one header label, producing an incomplete and structurally incorrect reconstruction."},{"id":"c66e2b14-a549-4ce7-bbd9-4f0fb50a2057","tool":"upstage-ai","tool_name":"Upstage AI","verdict":"failed","score":null,"score_total":null,"note":"Misaligns column headers with data regions in dense financial tables, producing a structurally inconsistent balance-sheet extraction rather than a clean header-to-row mapping."}]}