DeepSeek v3 0324 — Benchmarks
Benchmark scores for DeepSeek v3 0324 aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Overall rank: #47 of 123 open modelscomposite 52/100 across 9 benchmarks in 4 categories · methodology
Coding
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| Aider Polyglot | 55.1% | #8 / 18 | #33 / 69 |
| SciCode | 35.8% | #30 / 50 | #125 / 155 |
| SWE-bench Verified | 42.0% | #15 / 16 | #116 / 162 |
Knowledge
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| MMLU-Pro | 81.3% | #18 / 119 | #57 / 259 |
Math
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| AIME 2024/2025 | 37.8% | #44 / 74 | #190 / 270 |
| MATH Level 5 | 75.5% | #5 / 32 | #40 / 108 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| GPQA Diamond | 67.6% | #31 / 83 | #168 / 291 |
| SimpleBench | 27.2% | #18 / 23 | #84 / 101 |
| CritPt | 0.0% | #37 / 50 | #134 / 167 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.