Apex Shortlist
We show this table for reference; we do not rank on it.
A shortlist subset of the Apex mathematical reasoning benchmark reported in DeepSeek-V4 model evaluations.
Benchmark score on Apex Shortlist — October 29, 2026
We compile the Apex Shortlist rows from provider self-reports. DeepSeek V4 Pro 0813 leads the table at 90.2%, followed by Nemotron 3.5 Super VL 120B A12B NVFP4 (89.4%) and A.X K2 (88.6%). We do not use these results to rank models overall.
DeepSeek V4 Pro 0813
DeepSeek
Nemotron 3.5 Super VL 120B A12B NVFP4
NVIDIA
A.X K2
SK Telecom
9 modelsMathematicsCurrentDisplay onlyUpdated October 29, 2026
Benchmark score table (9 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | DeepSeek V4 Pro 0813DeepSeek | 90.2% | Not reported | Open |
| 2 | Nemotron 3.5 Super VL 120B A12B NVFP4NVIDIA | 89.4% | Not reported | Open |
| 3 | A.X K2SK Telecom | 88.6% | Not reported | Open |
| 4 | Nemotron 3.5 Super VL 120B A12BNVIDIA | 88.4% | Not reported | Open |
| 5 | DeepSeek V4 Flash 0731DeepSeek | 85.7% | Not reported | Open |
| 6 | DeepSeek V4 Pro (High)DeepSeek | 85.5% | Not reported | Open |
| 7 | DeepSeek V4 Flash (High)DeepSeek | 72.1% | Not reported | Open |
| 8 | DeepSeek V4 FlashDeepSeek | 9.3% | Not reported | Open |
| 9 | DeepSeek V4 ProDeepSeek | 9.2% | Not reported | Open |
Among the reported Apex Shortlist rows, DeepSeek V4 Pro 0813 is first at 90.2%. The third row is 1.6 points behind. The broader top-10 range is 81.0 points, so the table still separates the published systems.
9 models have been evaluated on Apex Shortlist. The benchmark falls in the Mathematics category. Apex Shortlist is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About Apex Shortlist
Year
2026
Tasks
Advanced mathematical reasoning
Format
Pass@1 math benchmark
Difficulty
Frontier math reasoning
BenchLM stores Apex Shortlist separately from the broader Apex row so provider-reported table values remain traceable.
Freshness and provenance
Version
Apex Shortlist 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 Apex Shortlist measure?
A shortlist subset of the Apex mathematical reasoning benchmark reported in DeepSeek-V4 model evaluations.
Which model scores highest on Apex Shortlist?
DeepSeek V4 Pro 0813 by DeepSeek currently leads with a score of 90.2% on Apex Shortlist.
How many models are evaluated on Apex Shortlist?
9 AI models have published results on Apex Shortlist 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.