Evidence · first-party tested

Across an 84-page hybrid financial report, the parser preserves heading order and section relationships so the extracted markdown stays aligned with the source structure.

✓ Worked🧾 artifact-verifiedTested Jun 20, 2026Tensorlake →
Output — unretouched
Target annual report page with growth-story summary.
Target annual report page with growth-story summary.
Output — unretouched
Parsed text hierarchy for the annual report page
Parsed text hierarchy for the annual report page
The observation

Across an 84-page hybrid financial report, the parser preserves heading order and section relationships so the extracted markdown stays aligned with the source structure.

Criterion: Reading Order & Structure · Scenario: Sumitomo Heavy Industries Consolidated Financial Report (group: financial-report)
Query this
get_evidence({
  tool: "tensorlake",
  scenario: "financial-report"
})
MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Same input, other tools
Adobe PDF Extract API ✗ FailedExtend AI ✓ WorkedLanding AI ✗ FailedLlamaparse ✓ WorkedMistral AI ✗ FailedNutrient.io ✓ WorkedPDF Vector ⚠ StruggledUpstage AI ✗ Failed
Real outputs, no retouching · every cell queryable via API & MCP · aidemos.com