Scores updated · 10 tests both models report · How we compare
Every test, side by side
All 10 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Other results10 tests
- HumanEvalDeepSeek-V3.1-Base by 24.472.548.1+24.4
- BBHDeepSeek-V3.1-Base by 2388.265.2+23
- MMLUDeepSeek-V3.1-Base by 17.487.470+17.4
- MATHDeepSeek-V3.1-Base by 17.362.645.3+17.3
- TriviaQADeepSeek-V3.1-Base by 17.283.566.3+17.2
- MMLU-ProDeepSeek-V3.1-Base by 16.458.842.4+16.4
- GSM8KDeepSeek-V3.1-Base by 1491.477.4+14
- C-EvalDeepSeek-V3.1-Base by 12.89077.2+12.8
- MBPPDeepSeek-V3.1-Base by 10.874.663.8+10.8
- CMMLUDeepSeek-V3.1-Base by 10.688.878.2+10.6
Questions people ask
How do you compare the two?
We use the 10 benchmark tests both models have published scores on. None of them is in the eight capability areas we count, so this page lists them without a verdict. Each score is the one shown on the model's own page.