Skip to content
K = 0.731
Benchmark sources

OfficeQA Pro

Databricks / Professional work · Retrieves information, calculates and answers questions from large collections of real financial documents.
Official benchmark
Across e/accUnder observation
Evidence budgetNot scored
Upstream data timeUnverified
Last successful syncCollection not started

What it measures and how

Fix Pro’s 133 tasks, separately from Full’s 246 and Pro V2’s 90. End-to-end retrieval and directly supplied answer pages are different conditions. Claude Agent SDK, Codex SDK and Antigravity CLI results belong to their runtime systems, not a common base-model benchmark. This measures document comprehension, not office-document output quality.

How this evidence is used

Under observation: the organization continues updates, but current results mainly compare different product runtime systems and the raw aggregation protocol needs verification. Retained as a document-comprehension reference; scores from different tools are not directly attributed to base models.

Limitations and attribution

Code and task licenses do not automatically cover every external result. Old and new task sets, end-to-end retrieval and directly supplied answer pages cannot be ranked together. Scores are not automatically collected.

Data license: Code is Apache 2.0. Task data is CC BY-SA 4.0 and has moved to a Hugging Face dataset requiring approved access.

Results are published by Databricks. Raw scores and the e/acc consensus index use different scales and cannot be added directly.