DeepSeek V3.2 Speciale
DeepSeek
Highlights
Top benchmark results for deepseek/deepseek-v3-2-speciale-2025-12-01.
96#1
2701#2
0.86#9
0.31#7
0.84#1
0.89#1
Benchmark table
| Benchmark | Category | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|
| AIME 2025 | math | 96 | - | Yes | Source |
| Codeforces | - | 2701 | - | Yes | Source |
| GPQA Diamond | general-knowledge | 0.86 | - | Yes | Source |
| Humanity's Last Exam | - | 0.31 | - | Yes | Source |
| IMO Answer Bench | - | 0.84 | - | Yes | Source |
| LiveCodeBench | - | 0.89 | - | Yes | Source |
Benchmark comparisons
Use the selector to switch benchmarks and see how this model stacks up against its closest competitors.
AIME 2025
Compare this model with the leading peers for the selected benchmark.
Benchmark
96
Rank #1/40
40 models
Showing 11 models around the selected model (out of 40 total).