SpreadsheetBench 2
We show this table for reference; we do not rank on it.
A spreadsheet-focused benchmark for agentic analysis and editing workflows.
Benchmark score on SpreadsheetBench 2 — November 13, 2026
We compile the SpreadsheetBench 2 rows from provider self-reports and secondary reports. ERNIE 5.5 Flash Preview leads the table at 36.4%, followed by LongCat-2.5 (34.9%) and Kimi K3 (34.8%). We do not use these results to rank models overall.
ERNIE 5.5 Flash Preview
Baidu
Reported by BaiduLongCat-2.5
Meituan
Reported by MeituanKimi K3
Moonshot AI
Reported by Moonshot AI4 modelsAgenticCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (4 models)
None of the 4 results on this page has a documented run by the benchmark or an independent evaluator: 3 reported by the labs that built the models and 1 reported in another publisher’s document.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | ERNIE 5.5 Flash PreviewBaiduReported by Baidu | 36.4% | Reported by Baidu | Not reported | Pending |
| 2 | LongCat-2.5MeituanReported by Meituan | 34.9% | Reported by Meituan | Not reported | Open |
| 3 | Kimi K3Moonshot AIReported by Moonshot AI | 34.8% | Reported by Moonshot AI | Not reported | Open |
| 4 | Ling 3.0 Flash FinInclusionAIReported by a third party | 21.8% | Reported by a third party | Not reported | Closed |
- Lab-reported
- Benchmark-run
- Independent
Among the reported SpreadsheetBench 2 rows, ERNIE 5.5 Flash Preview is first at 36.4%. The third row is 1.6 points behind. The broader top-10 range is 14.6 points, so the table still separates the published systems.
4 models have been evaluated on SpreadsheetBench 2. The benchmark falls in the Agentic category. SpreadsheetBench 2 is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About SpreadsheetBench 2
Year
2026
Tasks
Spreadsheet analysis and editing tasks
Format
Agent task-completion score
Difficulty
Professional spreadsheet work
Moonshot evaluates Kimi K3 with the Claude Code harness. BenchLM stores the provider-exact value as display-only productivity evidence.
Freshness and provenance
Version
SpreadsheetBench 2 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 SpreadsheetBench 2 measure?
A spreadsheet-focused benchmark for agentic analysis and editing workflows.
Which model scores highest on SpreadsheetBench 2?
ERNIE 5.5 Flash Preview by Baidu currently leads with a score of 36.4% on SpreadsheetBench 2.
How many models are evaluated on SpreadsheetBench 2?
4 AI models have published results on SpreadsheetBench 2 in the BenchLM catalog.
Compare top models on SpreadsheetBench 2
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 7,000+ subscribers.
One email each week. Unsubscribe anytime.