Skip to main content
Benchmark Registry
ModelsBenchmarksOrganizationsCompare

MMLU-Pro

Benchmark
MMLU-Pro
Evaluated by
Qwen Team
Release date
June 3, 2024
Version
MMLU-Pro Original
Metric
Accuracy

Results

6 results

LatestHistory
All providersAlibaba (Tongyi)DeepSeek AIGoogle (DeepMind)Microsoft AIMistral AIMoonshot AINVIDIAZ.ai
MMLU-Pro Original results
ProviderModelScoreSourceRegistry No.
DeepSeek AIDeepSeek-V4-Flash (non-thinking)83.0%Source for DeepSeek-V4-Flash (non-thinking) on MMLU-Pro Original (opens in a new tab)110001
DeepSeek AIDeepSeek-V4-Flash (max)86.2%Source for DeepSeek-V4-Flash (max) on MMLU-Pro Original (opens in a new tab)110001
DeepSeek AIDeepSeek-V4-Flash (high)86.4%Source for DeepSeek-V4-Flash (high) on MMLU-Pro Original (opens in a new tab)110001
DeepSeek AIDeepSeek-V4-Pro (max)87.5%Source for DeepSeek-V4-Pro (max) on MMLU-Pro Original (opens in a new tab)110002
DeepSeek AIDeepSeek-V4-Pro (high)87.1%Source for DeepSeek-V4-Pro (high) on MMLU-Pro Original (opens in a new tab)110002
DeepSeek AIDeepSeek-V4-Pro (non-thinking)82.9%Source for DeepSeek-V4-Pro (non-thinking) on MMLU-Pro Original (opens in a new tab)110002
Page 1 of 1

© 2026 Densa Labs

Toggle between the data update date and the application build time.
Legal
Color theme