Mathematics · Benchmark profile
A shortlist subset of the Apex mathematical reasoning benchmark reported in DeepSeek-V4 model evaluations.
Data verified 21 Jul 2026 · Methodology 1.6.0
Benchmark score on Apex Shortlist
DeepSeek V4 Pro (Max) leads at 90.2%, followed by DeepSeek V4 Flash (Max) (85.7%) and DeepSeek V4 Pro (High) (85.5%).
Visual analysis
Switch between model placement, score distribution and descriptive provider averages. Every view uses the same sourced leaderboard.
6 ranked models · higher is better
| Model | Rank | Provider | Score |
|---|---|---|---|
| DeepSeek V4 Pro (Max) | #1 | DeepSeek | 90.2% |
| DeepSeek V4 Flash (Max) | #2 | DeepSeek | 85.7% |
| DeepSeek V4 Pro (High) | #3 | DeepSeek | 85.5% |
| DeepSeek V4 Flash (High) | #4 | DeepSeek | 72.1% |
| DeepSeek V4 Flash | #5 | DeepSeek | 9.3% |
| DeepSeek V4 Pro | #6 | DeepSeek | 9.2% |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | DeepSeek V4 Pro (Max) deepseek-v4-pro-max | DeepSeek | open | Estimated reference | 90.2% |
| #2 | DeepSeek V4 Flash (Max) deepseek-v4-flash-max | DeepSeek | open | Estimated reference | 85.7% |
| #3 | DeepSeek V4 Pro (High) deepseek-v4-pro-high | DeepSeek | open | Estimated reference | 85.5% |
| #4 | DeepSeek V4 Flash (High) deepseek-v4-flash-high | DeepSeek | open | Estimated reference | 72.1% |
| #5 | DeepSeek V4 Flash deepseek-v4-flash | DeepSeek | open | Estimated reference | 9.3% |
| #6 | DeepSeek V4 Pro deepseek-v4-pro | DeepSeek | open | Reference only | 9.2% |
The top of this snapshot is led by DeepSeek V4 Pro (Max) at 90.2%; third place is 4.7 points behind. The top-6 spread is 81.0 points.
About Apex Shortlist
A shortlist subset of the Apex mathematical reasoning benchmark reported in DeepSeek-V4 model evaluations. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
A shortlist subset of the Apex mathematical reasoning benchmark reported in DeepSeek-V4 model evaluations.
DeepSeek V4 Pro (Max) by DeepSeek currently leads with 90.2%.
6 models in the LuminaBench cohort have a qualifying score on this benchmark.
Related
