Scores updated · 21 tests both models report · How we compare
Every test, side by side
All 21 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Other results21 tests
- SimpleQADeepSeek-V4-Pro-Base by 14.652.137.5+14.6
- Social IQAHy4-flash-base by 773.180.1+7
- MBPP+ (EvalPlus-augmented)DeepSeek-V4-Pro-Base by 5.479.173.7+5.4
- U-MATHHy4-flash-base by 564.869.8+5
- OCW CoursesDeepSeek-V4-Pro-Base by 4.268.264+4.2
- MMLU-ReduxDeepSeek-V4-Pro-Base by 3.290.887.6+3.2
- SuperGPQAHy4-flash-base by 354.857.8+3
- AGIEval-EnDeepSeek-V4-Pro-Base by 2.878.275.4+2.8
- HellaSwagDeepSeek-V4-Pro-Base by 2.88885.2+2.8
- MMLU-ProDeepSeek-V4-Pro-Base by 2.473.571.1+2.4
- ARC-ChallengeDeepSeek-V4-Pro-Base by 29795+2
- Chinese SimpleQA (C-SimpleQA)DeepSeek-V4-Pro-Base by 276.674.6+2
- IncludeDeepSeek-V4-Pro-Base by 1.584.382.8+1.5
- Belebele (5-language subset)DeepSeek-V4-Pro-Base by 1.495.293.8+1.4
- MMLUDeepSeek-V4-Pro-Base by 1.490.188.7+1.4
- C-EvalDeepSeek-V4-Pro-Base by 1.193.192+1.1
- C3DeepSeek-V4-Pro-Base by 178.377.3+1
- AGIEval (Chinese)tie89.890.6tie
- CMMLUtie90.891.6tie
- KOR-Benchtie55.454.7tie
- PIQAtie84.984.4tie
Questions people ask
How do you compare the two?
We use the 21 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.