OfficeQA Pro v2
We show this table for reference; we do not rank on it.
The second version of OfficeQA Pro: grounded reasoning over real financial and office documents through retrieval, cross-document verification, and multi-step numerical reasoning.
Benchmark score on OfficeQA Pro v2 — October 29, 2026
We compile the OfficeQA Pro v2 rows from provider self-reports. Kimi K3.1 Pro leads the table at 64.4%. We do not use these results to rank models overall.
1 modelMultimodal & GroundedCurrentDisplay onlyUpdated October 29, 2026
Benchmark score table (1 model)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Kimi K3.1 ProMoonshot AI | 64.4% | Not reported | Pending |
About OfficeQA Pro v2
Year
2026
Tasks
Document and spreadsheet tasks
Format
Task accuracy
Difficulty
Enterprise document reasoning
Moonshot evaluates with Claude Code (Codex for GPT models), renders every PDF in the corpus as images with no machine-readable text, and scores zero for tasks where a model retrieved reference answers online. BenchLM keeps v2 separate from the original OfficeQA Pro key.
Freshness and provenance
Version
OfficeQA Pro v2 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does OfficeQA Pro v2 measure?
The second version of OfficeQA Pro: grounded reasoning over real financial and office documents through retrieval, cross-document verification, and multi-step numerical reasoning.
Which model scores highest on OfficeQA Pro v2?
Kimi K3.1 Pro by Moonshot AI currently leads with a score of 64.4% on OfficeQA Pro v2.
How many models are evaluated on OfficeQA Pro v2?
1 AI models have published results on OfficeQA Pro v2 in the BenchLM catalog.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.